Oscilloscopes to stay away from? (spec: Tek 2337 question)

[Jschool]

Dave suggests you buy an old analog oscilloscope off of eBay, sticking with the major brands.  Easy enough.

But I've noticed some of the ones I find more desirable, like the Tektronix 2337, that upon reading some experiences from users it seems they're more difficult to work on or repair because they used a lot of custom parts.  And that the 475 and similar models used standardized parts.  Is there anything to this?

I really like the 2337 for its size and features, but I don't want to sink a lot of money into a beginner scope that can't be easily fixed either.

[Fraser]

Peronally I don't recommend a Tektronix CRO for beginners unless it's very cheap. Don't get me wrong, they are superb scopes and are mostly reliable. The problem is many cheap ones are now 20+ years old and the odd failure is to be expected. In terms of repair the Tektronix oscilloscopes are a significant challenge as some of the custom hybrids and other parts are both rare and expensive. EHT modules are not an uncommon failure and can be hard to repair. More sophisticated Tektronix scopes are gorgeous but what do you do when a custom Hybrid or CCD dies ? Such failures are not uncommon amongst ebay auction units and that is often why they are sold on. Some fault symptoms can be hard to spot unless you really know your stuff, such as CCD failures....yes some use Charge Coupled Devices in their designs and not as cameras  

For the beginner or someone who wants a relatively easily repaired oscilloscope with good manufacturer support, I would recommend HAMEG CRO's. Not necessarily the latest models though. As a beginners CRO with 20MHz BW you won't go far wrong with one of the following:

HM203-x Beginner/basic functionality
HM204-x Intermediate functionality inc delayed sweep
HM205-x Storage CRO

The 30 x and 50x ranges are more pricey and the most recent models can have reliability issues with their EHT and focus circuits. HAMEG are now owned by R&S.


Many of the HAMEG range contained a nice component tester that works and isn't a gimmick.

The components used are nice and easy to find so repair will be possible for some time to come. Obviously tube failure would be a bummer but is usually rare except for drop damage. You can download the manuals and schematics for free from the Hameg site.

Do not pay too much for a HAMEG.... GBP50 will usually secure a decent one and many HM203 models sell for much less (~GBP20) or even free if you find someone having a clearout.

Oh, and for transparency.....I used to sell these when I was a student :-) Great customer support from Germany  

[DaveW]

As a beginners CRO with 20MHz BW you won't go far wrong with one.

i keep hearing this, but i dont get this. if you go advance into things like FPGA, 20MHz is not a good thing at all! worst if you are keen into radio. i'm still thinking of how to crank out 400MHz effectively out of my 50MHz DS1052E. buy as much bandwidth as you can possibly afford! i agree with the rest of points though on repairing obsolete stuffs.

There are points where you need high amounts of bandwidth for debugging FPGA applications. However, most of the time you're looking at high speed data signals so a high speed logic analyser is the correct tool for the job. Similarly for radio, there are few times that you actually need to look at the signal output. Most of the time you are looking at the mixer inputs and output at the IF, so for most radio uses a 50MHz scope is plenty and 20MHz is enough for most applications.

It seems to be mostly a matter of finding the right age scope. Not so old that it's coming up to failure, and not so new that it's unrepairable! At least with the older Tektronix stuff there's a large online community that can help with fixing. Although as you say the Hameg support is excellent, always happy to provide service manuals...

[tekfan]

Well the Tek 2300 series are great if they work. Small and very rugged. The ones that still work probably will keep on going for a long time. This series has specially high performance power FETs selected for the vertical amplifier. If they fail and you replaced them with a similar one, the max bandwidth would be seriously degraded. The low frequency signals would be unaffected. But I think if you looked hard enough you could find the correct FETs. They have a slightly reduced CRT (not the standard 10x8cm type)


The 400 series is probably the most well known for it's reliability, ruggedness (all metal case), etc.
465a/b - 100MHz, dual trace
466 - 100MHz, analog storage, dual trace
468 - 100MHz, digital storage, dual trace, RS-232 output
475 - 200MHz, dual trace, 2mV/div sensitivity
485 - 350MHz, dual trace, internal 50ohm termination (quite a beast but still somewhat portable)

All the models have delayed time bases

You can't get anything wrong with buying anyone of these. They are very easy to repair, have linear supplies (none of that noisy switching stuff). The 465 now almost holds a legendary statusamong Tek scopes. It's been the world's standard delayed sweep dual trace scope for nearly 30 years.

2200 series are a bit more flimsy than the 400 series since their front panel is plastic (same as the 2300 series)

2212 - 60MHz combined analog/digital storage scope
2230 - 100MHz, same as above
2232 - 100MHz, same as above but a bit more advanced with softkeys
2235 - 100MHz, analog, general porpouse, basically a revised and lightweight 465
2236 - 100MHz, same as 2235 but with added frequency counter, timer, multimeter

All of them have 2mV/div sensitivity, delayed time base, all the units weigh as little as 8 kilos

A couple of years ago I bought the 2236 on American ebay. I live in Europe so the shipping and customs cost almost as much as the scope. But it was worth every single penny. The really nice feature of the scope is the counter. Press a button and you get a 8 digit readout of the frequency. It also does period and time difference measurments, gated frequency or period measurments of short bursts. The scope even succesfuly triggerers on signals at almost 150MHz.
If you want to measure the voltage of channel 1 just select CH1 volts and set the coupling switch to DC or AC, whichever you would like to measure. The AC range is even true RMS and works to about 200KHz. On the side of the scope are two jacks for standard shrouded multimeter probes. With the press of a button you can measure again DC or AC (autoranging 5000 count), or you can measure ohms to 2 Gohms. No milliamp range but what else do you want for a 30 year old scope!

This isn't a review but I just want to help people to decide which scope to buy.

Repair of the 2300 series scopes is also not a problem. Most of the ICs are standard off the shelf parts apart for some of the EPROMS. Even the time base switch was originally used in ham radios. All the electronics are found on a couple of big boards. The design philosophy was to put all the circuitry on a single board and get rid of the unreliable connectors and ribbon cables.

Probably the only thing that can go wrong in one of these is the switching power supply. Again standard parts and easily repairable.
The CRT is probably one of the more reliable parts of any scope (that is if you don't throw it down the stairs).

A bit more on the Hameg scopes
We have a lot of HM303s (35MHz). Sadly they haven't stood the test of time. They were bought new somewhere in the 90's. One third of them developed problems. Either reduced bandwidth, bad attenuator contacts, weird oscillations of the time base.
The 205 series is much better. We've never had any problems with these. They are nice scopes but are higher in price on the second hand market. The only drawback is the low bandwidth.

[Fraser]

Mechatrommer,

Regret FPGA's etc are well away from my area of experience and I agree that a 20MHz CRO would not suffice for the high frequency stuff.....but then what bandwidth CRO would ? With 500MHz square waves on my PC's bus I would need one heck of an expensive CRO to cope with that .....I don't think CRO's are the best tool at those higher frequencies  

Now as for RF at 400MHz...you are now talking my language  

I carry out repair and development work up to 1.7 GHz and some ISM band 2.4GHz and K/Ku SHF stuff now and again. When carrying out diagnostics at the full operating frequency, I would not be looking for any form of oscilloscope as my primary tool. If accurate level measurement is needed, I use a True RMS RF Millivoltmeter or Power meter. If frequency is important, I use a UHF or SHF frequency counter with GPS disciplined oscillator reference. And finaly, for linearity and filtering tests I would reach for my Advantest R4131D 3.5GHz spectrum anlayser and hunt for the tell tales of distortion spurii, harmonics and filter leakage. I would only use an analogue CRO when looking at the lower frequencies in the IF stages of the RX or TX and even then a spectrum analyser can be more useful. For reciever testing and repair it is often possible to work at the lower IF frequencies in order to diagnose issues in preceeding RF stages. You are effectively using the receivers own RF downconversion for diagnostic purposes so a CRO would be usable.

The bad news is that the cost of a DS1052E pales into insignificance when looking to buy RF test kit capable of 3GHz working. My Advantest R4131D originally cost GBP15,000 and, 10 years later, now costs around GBP2,000 t0 GBP3,000 on the used market. I can see why you would like to use the DS1052E for the role at 400MHz but considering it's limited bandwidth, sampling rate and the fact that it is a budget DSO, regret that it is definitely not the tool for the job.

[cs.dk]

Well the Tek 2300 series are great if they work. Small and very rugged. The ones that still work probably will keep on going for a long time. This series has specially high performance power FETs selected for the vertical amplifier. If they fail and you replaced them with a similar one, the max bandwidth would be seriously degraded. The low frequency signals would be unaffected. But I think if you looked hard enough you could find the correct FETs. They have a slightly reduced CRT (not the standard 10x8cm type)


The 400 series is probably the most well known for it's reliability, ruggedness (all metal case), etc.
465a/b - 100MHz, dual trace
466 - 100MHz, analog storage, dual trace
468 - 100MHz, digital storage, dual trace, RS-232 output
475 - 200MHz, dual trace, 2mV/div sensitivity
485 - 350MHz, dual trace, internal 50ohm termination (quite a beast but still somewhat portable)

All the models have delayed time bases

You can't get anything wrong with buying anyone of these. They are very easy to repair, have linear supplies (none of that noisy switching stuff). The 465 now almost holds a legendary statusamong Tek scopes. It's been the world's standard delayed sweep dual trace scope for nearly 30 years.

2200 series are a bit more flimsy than the 400 series since their front panel is plastic (same as the 2300 series)

2212 - 60MHz combined analog/digital storage scope
2230 - 100MHz, same as above
2232 - 100MHz, same as above but a bit more advanced with softkeys
2235 - 100MHz, analog, general porpouse, basically a revised and lightweight 465
2236 - 100MHz, same as 2235 but with added frequency counter, timer, multimeter

All of them have 2mV/div sensitivity, delayed time base, all the units weigh as little as 8 kilos

A couple of years ago I bought the 2236 on American ebay. I live in Europe so the shipping and customs cost almost as much as the scope. But it was worth every single penny. The really nice feature of the scope is the counter. Press a button and you get a 8 digit readout of the frequency. It also does period and time difference measurments, gated frequency or period measurments of short bursts. The scope even succesfuly triggerers on signals at almost 150MHz.
If you want to measure the voltage of channel 1 just select CH1 volts and set the coupling switch to DC or AC, whichever you would like to measure. The AC range is even true RMS and works to about 200KHz. On the side of the scope are two jacks for standard shrouded multimeter probes. With the press of a button you can measure again DC or AC (autoranging 5000 count), or you can measure ohms to 2 Gohms. No milliamp range but what else do you want for a 30 year old scope!

This isn't a review but I just want to help people to decide which scope to buy.

Repair of the 2300 series scopes is also not a problem. Most of the ICs are standard off the shelf parts apart for some of the EPROMS. Even the time base switch was originally used in ham radios. All the electronics are found on a couple of big boards. The design philosophy was to put all the circuitry on a single board and get rid of the unreliable connectors and ribbon cables.

Probably the only thing that can go wrong in one of these is the switching power supply. Again standard parts and easily repairable.
The CRT is probably one of the more reliable parts of any scope (that is if you don't throw it down the stairs).

A bit more on the Hameg scopes
We have a lot of HM303s (35MHz). Sadly they haven't stood the test of time. They were bought new somewhere in the 90's. One third of them developed problems. Either reduced bandwidth, bad attenuator contacts, weird oscillations of the time base.
The 205 series is much better. We've never had any problems with these. They are nice scopes but are higher in price on the second hand market. The only drawback is the low bandwidth.

Great post!!

I'm also looking for something to begin with..
What can you tell about the 24xx-series?

[tekfan]

The 2400 are large format portable scopes. All of them have dual delayed time bases all the analog ones have 4 channels. 2 have standard attenuators in 1-2-5 sequence. Sensitivity is from 5V/div down to 2mV/div. The other two channels are selectable 100mV/div and 5mV/div (probably the most usefull ranges). These are excluded in the digital scopes. All of the scopes feature (unlesso therwise noted) readouts, cursor functions and various options are (were) available including integrated frequency counter/timer, GPIB, setting storage

A few notes about repairing these:
These all use switching power supplies. Usually the only problem is sometimes the capacitors in the supply start to have high ESR which usually leads to noisy supply lines or the scope not even turning on. This is easy to repair.
The other thing is these scopes use hybrid ICs which can develop corroded contacts. These can also be cleaned and the problem goes away. A sign of corroded contacts of the ICs are failed self tests. Yes these scopes have a self test facility which aids in identifying which part of the scope needs repair. It is a very small chance of them failing but if one of them does they can regularly be seen on ebay. Going from anywhere from 10-50 EUR. A small price to pay for such an advanced scope.

No doubt these are one of the most advanced analog scopes ever built. Only surpassed by the Tek 11000 series analog scopes (much much much bigger but still worth every penny).

On with the list:
The large format portable scopes actually start from the high end 2200 series:
2245 - 100MHz, 4 channel, analog, basic large format scope, no cursors, no setting storage
2246 - 100MHz, 4 channel, analog, advanced triggering (TV line and field trig)
2247 - 100MHz, 4 channel, analog, advanced triggering (TV line and field trig), added function with high resolution counter/timer

2430 - 150MHz, 100MS/s, 2 channel, digital storage, only limited trigger functionality
2440 - 300MHz, 0.5GS/s, 2 channel, digital storage,  other same as 2430
2445 - 150MHz, 4 channel, analog, internal 50 ohm termination, available high resolution counter/timer, optional TV(line, field) trigger
2465 - 300MHz, otherwise same as 2445
2465A - 350MHz, otherwise same as 2445
2465B - 400MHz, otherwise same as 2445
2465 DMS - 400MHz, added multimeter on top of scope, otherwise same as 2445
2465 CTS - 400MHz, high resolution counter/timer as standard, otherwise same as 2445
2465 DVS - probably the same thing as the 2465 DMS

In a class of it's own:
2467B - 400MHz, otherwise same as 2445 plus MCP (microchannel plate) inside the CRT which allows a single 1ns pulse to be viewed in normal roomlight
2467 BHD - 400MHz, specially designed for development and work on high definition broadcast and television, TV trigger standard of course, otherwise same as 2467B, optional triggering from logic probe (no logic analyzer functions), otherwise same as 2467B


Like most Tek products, the bandwidth ratings are very conservative. Most scopes will almost have double the rated bandwidth (amazing, I know).
To prove it, here is a picture of a fast pulse viewed by a 2445A 150MHz scope, the rise time is 1.24ns which equates to a bandwidth of more than 270MHz

I have a suspicion this is the Jim Williams avalanche pulse generator that's being fed into the scope

Anyway, whichever scope you buy from this series surely won't let you down.
I don't currently own one from this list as there is not enough space at my bench.

[slburris]

Don't at least some of the 4XX series use tunnel diodes for their triggering
circuits?  TDs are unobtainium except from donor scopes, although I've heard
of some success people have had with Russian TDs obtained on Ebay.

Personally, I have the 2236, 2246, and 2465 scopes and love them all.

No one has mentioned the 7000 series, which take plugin modules and at
least one can go up to 1Ghz, the 7104 I believe.

There are some "fake" 2465 scopes out there, which are 2445's with some bandwidth
limiting parts clipped out.  There was a discussion about this on the Yahoo tekscopes list
a few months ago.

Scott

[tecman]

right! $300 is something impossible for RF work by today's standard. i'll keep looking for a spectrum analyzer that can fit into my pocket money... none so far, new or used

I was also searching.  Eventually got a Tek 2714 for $ 600 on eBay, advertised as in good shape.  It is designed for CATV, but you can menu out of the cable related menus and you have a 9 KHz to 1.8 GHz general purpose analyzer.  After I received it, I found it would not respond to any key entries.  Next challenge was opening it.  It had been slightly dented in the corner of the case, but since the case is a glove tight fit, it would not slide off.  After several hours of prying, pushing and pulling I finally got the case off.  I downloaded the 2712 manual (sister of the 2714 except for the menus) from the Tek site.  A bit of troubleshooting  pointed to a 4538 dual cmos one-shot.  Replaced the 4538 and all worked.  I then carefully hammered the dent out of the cover.  The last change I made was to change the input connector.  The 2714 has a threaded front panel connector that mates with a CATV F connector and is 75 ohm input Z.  Looking at the schematics, the input attenuator is a 50 ohm attenuator, and the input jack assembly has a series 25 ohm resister internally for the 75 ohm input.  I ordered a new BNC bulkhead connector (they have an SMA on the back) which fit like the old one, just requiring a slight enlargement of the front panel opening.  An amplitude recalibration to compensate for the 25 ohm loss in the old connector, and now it is a great general purpose 50 ohm analyzer.

paul

[Fraser]

Off topic but applicable...

I have a Tektronix 2710 1.8GHz Spectrum Analyser languishing under my lab bench. It's a very nice piece of kit but sadly it started to have display text issues where it would fail to display some digit locations. A little later it's PSU went BANG ! and the main swtching transistor blew to pieces. All new low ESR capacitors, switching controller IC, switching transistor, switching drive transistor and inverter transistors fitted. I also replaced some SMD transistors that had been potentially stressed by the original failure. All tested Ok around the PSU with power off.

Power was applied and BANG! goes another switching transistor and all associted components that were previously replaced. I am used the the foibles of SMPSU's but this one has tested my patience to date and I know there is still a fault in that PSU but damned if I can find it at the moment  The unit has been set aside until I have more time. I already have a staged diagnostic procedure in mind to isolate parts of the unit that could cause such a failure.

The lesson....whilst TEK equipment is gorgeous it can be a pain to repair and even when I fix the PSU I know I have a problem with the display board that is very likely unrepairable from comments I have read regarding the likely failed component.   

Where beginners are concerned I firmly believe the KISS principle applies to test kit, to ensure it is maintainable.

[mikeselectricstuff]

The 2465A was probably the best general-purpose analogue scope ever made. The main issue is that they used some custom ICs that were unobtanium - they don't fail often but if they do it's Game Over. However the comments above from Tekfan suggest that now that analogue scopes are much less in demand and prices have dropped, there may now be enough recycled spares floating around at reasonable prices, but the supply will dwindle eventually
The moral is don't buy a dead one unless it's cheap enough to be a gamble - working ones seem to be available for around the price of a much lower-end digital scope these days.
You need to pay for a DPO/invinivision type digital scope to get comparable performance to something like a 2465.

[Teknotronix]

The 2465A was probably the best general-purpose analogue scope ever made. The main issue is that they used some custom ICs that were unobtanium - they don't fail often but if they do it's Game Over. However the comments above from Tekfan suggest that now that analogue scopes are much less in demand and prices have dropped, there may now be enough recycled spares floating around at reasonable prices, but the supply will dwindle eventually
The moral is don't buy a dead one unless it's cheap enough to be a gamble - working ones seem to be available for around the price of a much lower-end digital scope these days.
You need to pay for a DPO/invinivision type digital scope to get comparable performance to something like a 2465.

Has anyone reverse engineered the IC's or have a block diagram?

[EEVblog]

The 2465A was probably the best general-purpose analogue scope ever made.

I'd concur with that. They are still highly sort after on ebay.
Don't forget the 2225 is you are doing low level signal work. The 500uV/div sensitivity is pretty much unmatched by any other scope?
http://republika.pl/rubi/oscyloskopy/tek/t.pdf

Dave.

[mikeselectricstuff]

The 2465A was probably the best general-purpose analogue scope ever made. The main issue is that they used some custom ICs that were unobtanium - they don't fail often but if they do it's Game Over. However the comments above from Tekfan suggest that now that analogue scopes are much less in demand and prices have dropped, there may now be enough recycled spares floating around at reasonable prices, but the supply will dwindle eventually
The moral is don't buy a dead one unless it's cheap enough to be a gamble - working ones seem to be available for around the price of a much lower-end digital scope these days.
You need to pay for a DPO/invinivision type digital scope to get comparable performance to something like a 2465.

Has anyone reverse engineered the IC's or have a block diagram?

Around 10-15 years ago, when 2465's were still fetching £2-3000  I did wonder why, with  off-the-shelf semiconductors 20-odd years advanced from where they were  when the scope as released, nobody had done PCBs to replace the hybrid modules. I do recall reading about 5 years later that someone was working on something, but by then digital scopes had caused the prices of analogue scopes to plummet, and a lot of the old 2465's with dead hybrids had probably been long since scrapped.

The 2465A was probably the best general-purpose analogue scope ever made.

I'd concur with that. They are still highly sort sought after on ebay.

A quick look on ebay shows 2465A's with  cal & warranty can be had between £300-500, which is amazing value for money.  When I got my MSO6000, I considered selling my 2465A, but the prices they fetched just weren't worth it.
 

[Sigmoid]

What's the say on Tek 2201s? I saw one with a seller warranty on eBay, but quite far from cheap for a 20Mhz scope.
How complex are they from the inside (hybrids, etc.)? Is it likely to develop problems?

[robrenz]

Don't forget the 2225 is you are doing low level signal work. The 500uV/div sensitivity is pretty much unmatched by any other scope?

Except a 7000 series with a 7A22 differential amp plugin at 10µV/div

[vk6zgo]

Don't forget the 2225 is you are doing low level signal work. The 500uV/div sensitivity is pretty much unmatched by any other scope?

Except a 7000 series with a 7A22 differential amp plugin at 10µV/div

Yeah,the 7000 series are nice old beasts--you can get such a range of plugins.

I'm looking for a 7L12 or similar Spectrum Analyser plugin.
Rotten thing is,I saw several on eBay Aust before I got the 7613,but nary a one,since!

[Sigmoid]

I've been thinking about it for about a minute, but I dropped the idea. They are just too huge, and I'm a bit afraid of buying something this old that I cannot really use to its full capabilities.
Currently I'm hesitating between a 2201 (seller refurbished, 30 days warranty, but seems expensive for $150), and a tested 2225 in really good shape (if I can score it)... I've been considering a Hameg given how I have a history with them, but those vernier pots (and all the mechanical wizardry with the front panel) seem really easy to break and hard to fix.

[kfitch42]

I bought a 2336YA (nearly identical to the 2337) on epay a while back. $50 including shipping. The pics showed it displaying a flat line, turns out the coax going to the input bncs had fallen off. And it was missing a button (trigger view), a chop stick and some duck tape fixed that .

After I got it (and fixed those two simple problems) I realized there was no freely available service manual. I spent $10 to get an electronic copy of one. So far I have liked it pretty well, though I haven't used it a ton. It is pretty far out of cal, but one day I will actually attempt to get it close-ish to calibrated now that I have the service manual.

The good and bad is that the scope is small and portable. Good being portable. Bad being that things are a bit more packed in when trying to service it.

[Sigmoid]

Lol it's one of these military ones, right? Looks pretty cool, but I think I'll stay with the civilian models.

I bought a 2336YA (nearly identical to the 2337) on epay a while back. $50 including shipping. The pics showed it displaying a flat line, turns out the coax going to the input bncs had fallen off. And it was missing a button (trigger view), a chop stick and some duck tape fixed that .

[GK]

Don't forget the 2225 is you are doing low level signal work. The 500uV/div sensitivity is pretty much unmatched by any other scope?

Except a 7000 series with a 7A22 differential amp plugin at 10µV/div

Yeah,the 7000 series are nice old beasts--you can get such a range of plugins.


I'm looking for a 7L12 or similar Spectrum Analyser plugin.
Rotten thing is,I saw several on eBay Aust before I got the 7613,but nary a one,since!

My much older 1A7 plug-in for 500 series scopes does 10uV/division with DC coupling. Parallel push-pull nuvistor input stage. DC drifts like fark though.

[Sigmoid]

By the way, does anyone know a good online reference with release dates, manufacturing history, specs and possibly some words about the major Tektronix models? (Like the differences between a 2336 and a 2337, etc.)

[alm]

There's also a plugin for the Tek 5000 series and even a stand-alone TM-500 version (AM502) of this plugin. Probably all based on a very similar design (I know for a fact that the 7000/5000/TM500 were very similar). So while 500 uV/div is decent for a portable scope, it's definitely been matched and exceeded by lab scopes.

[GK]

By the way, does anyone know a good online reference with release dates, manufacturing history, specs and possibly some words about the major Tektronix models? (Like the differences between a 2336 and a 2337, etc.)

http://w140.com/tekwiki/wiki/Main_Page

[PaulAm]

The 24xx list didn't mention the 2432/2432a DSO. which is inbetween the 2430 and 2440 in performance  It's not a bad scope.

The 468 had a GPIB option.  The DSO section is only good to 10MHz though.

On the 2465, U800 is the primary problem.  There is a design with smd parts that's available that more or less is a drop in replacement.  I don't know first hand anyone who's tried it though.  Some people have had some success running it through a baking process.  It comes up a lot on the Tekscopes list.