Hacking the HDO1k/HDO4k Rigol 12 bit scope

[markone]

Interesting that they have put all 9 FPGA versions in that folder. Wonder why ...
For 50 ohm, even that the switch can be "moved", that need to be compensated as a value, along the signal path.

Why do you have to compensate it ?

It's not a signal source.

[skander36]

If is just a relay that route the signal to a passive limiter (impedance adapter), it will be lower as amplitude.

[markone]

If is just a relay that route the signal to a passive limiter (impedance adapter), it will be lower as amplitude.

So it's not a simple 50 ohms resistor placed in parallel to the input ?

[markone]

If you mean something like this, i guess the same I/O that switches 50 ohms will act on compensation attenuator bypass.

[Kleinstein]

The 50 Ohm in put is not with a resistgor in to ground at the input. There are different signal paths to the front-end chips.
It very much looks like the lower relay switched between 50 Ohm and 1 M ohm mode.
The upper relay looks like it switches between 2 attenuators for the 50 Ohm version and likely also for the 1 M version.
The 50 ohm paths are to the left side, with 2 clearliy visible dividers in the 50 ohm range, going to 2 pins of the front end chip.
Between the 2 realys there is also some attenuator that is too low in impedance to be used with the 1 M signal.

The 50 Ohm termination is thus likely split over multiple resistors: the 3 between the realys, one of the 2 attenuators on the left and to a small part the termination at the front end chip.

The 50 ohm front end looks essentially the same, so it should really be only a SW question of enabling a suitable inputs at the front end chip.
In the high frequency range the 50 ohm part may have a subtile effect on the 1 M path - so instead of redicigning the PCB they may have decided to populate the extra realy and few resistors for the 50 ohm part too.

There may however be a catch:  It looks like that in the HDO4000 there are 2 differential signals from each front end to the ADCs. With the HDO1000 only 1 differential signal per front end goes to the ADC.
So the 50 ohm path may just be a dead end at the output of the front end chips with no free ADC inputs to go to. Even the developers at Rigol may not have an easy way to add the 50 ohm option for the 4 channel 1 ADC combination.   With the 2 channel version there would be enough inputs.

[skander36]

Thank you both for the informations!
Looking to the block diagram, it seem that the frontend chip contain two paralell active stages and switch between them.
If it's correct, and not just an an abstract representation ...

[markone]

The 50 Ohm in put is not with a resistgor in to ground at the input. There are different signal paths to the front-end chips.
It very much looks like the lower relay switched between 50 Ohm and 1 M ohm mode.
The upper relay looks like it switches between 2 attenuators for the 50 Ohm version and likely also for the 1 M version.
The 50 ohm paths are to the left side, with 2 clearliy visible dividers in the 50 ohm range, going to 2 pins of the front end chip.
Between the 2 realys there is also some attenuator that is too low in impedance to be used with the 1 M signal.

The 50 Ohm termination is thus likely split over multiple resistors: the 3 between the realys, one of the 2 attenuators on the left and to a small part the termination at the front end chip.

The 50 ohm front end looks essentially the same, so it should really be only a SW question of enabling a suitable inputs at the front end chip.
In the high frequency range the 50 ohm part may have a subtile effect on the 1 M path - so instead of redicigning the PCB they may have decided to populate the extra realy and few resistors for the 50 ohm part too.

There may however be a catch:  It looks like that in the HDO4000 there are 2 differential signals from each front end to the ADCs. With the HDO1000 only 1 differential signal per front end goes to the ADC.
So the 50 ohm path may just be a dead end at the output of the front end chips with no free ADC inputs to go to. Even the developers at Rigol may not have an easy way to add the 50 ohm option for the 4 channel 1 ADC combination.   With the 2 channel version there would be enough inputs.

I do not understand the need to have two separate paths to ADC for this purpose, is not the front-end chipset in charge to select the different input (switching also related relays) acting as a MUX and then source the ADC ?

[Kleinstein]

The part where the 2 front end versions differ a little does not looks like it is a filter, it more looks like a compensation network to fine tweek the frequency response of the dividers.

The 50 Ohm paths are quite clear: first a Pi type attenuator with 294 Ohm + 17.8 Ohm + 294 ohm  and than the relay to chose between
a) another such attenuator  and than to the front end chip
b) an Attenuator with 56.2 Ohm + 383 ohm + 56.2 and than to another input of the chip

The 1 M path is a little less clear:
 frist a RC combination including on of the slightly strange looking part with a blue dot (this could be a laser trimmalbe resistor)
 the brown capacitor right of the relay for AC coupling and the photomos switch in parallel for DC coupling.
 Than to the upper switch of the relay and from there 2 paths:
a)   an AC coupled path to the chip, with again what looks like a trimmable correction network.
      in addition a divider with 681 K and  316 K + a small resistor for a DC path that goes to a seprate pin at the chip

b) correction network  +  750 K  with parallel capacitor (rel. large form factor by likely just some 10 pF)
     than lots of the small resistors and capacitors to at least 3 inputs of the chip. So like 3 different DC coupled attenuator steps.
    this inludes the small corner that changes between the versions.

I consider it very unlikely that the changed layout part in would effect the path with lower attenuation.

I don't even see any hardware 20 MHz or similar low pass filter. The parts with the possibly trimmable resistors (with blue/orange dot) are more like effecting only a small part of the signal.
These can not be the BW limiting fitlers but more like fine corrections for a flat response.
So switchable BW limiting would be fully inside the front end chips or possibly even in software.


A possible weak spot could be that I don't see very mich protection for the 50 ohm mode. With the small attenuator there seem to be only 2 x 17.8 Ohms between the input and the front end chip.
So one should take the maxiumum voltage limit for the 50 ohm case serious. Ideally the relay may switch off the 50 Ohm mode if overlaod, but I would not count on this.

The front end chip definitely chooses between the different inputs. It looks like 2 x for 50 Ohm and 4 x for 1 M mode + an additional DC input that also has to be somehow combined with the AC part !
Likely there are 2 separate amplifier for the 1 M and 50 Ohm part - at least the 50 ohm input seem to be terminated inside and it would make sense to be lower noise here.
The question is how the switching is done after the amplifiers - they could combine the ouputs to one ouput or they could as well decide to keep them seprate
The front end chips have 2 outputs each. In the HDO1000 one 1 is used the others are missing the termination.
if they really use the same PCB, the unused outputs would likely lead to the unpopulated ADC. So there would not be the classic like of ADC1 for channels 1 and 2 and ADC2 for channels 3 and 4, but all 4 channels have a path to both ADCs. In this case the 50 Ohm mode much go the same path.

[skander36]

Also there can be differences on the back side, but I can't find a photo for 1K, only for 4K, from Dave's Flicker account.

[markone]

-snip
The front end chip definitely chooses between the different inputs. It looks like 2 x for 50 Ohm and 4 x for 1 M mode + an additional DC input that also has to be somehow combined with the AC part !
-snip

So the simple answer to my question is : Yes, we do not need a double path to ADC to provide 50 ohms input capability.

[markone]

There is another possible input layout,  used nowadays by some high performance instruments :

[Kleinstein]

Much of the external circuit is clear. The code 1B for a SOT23 part on the bottom side indicates a simple NPN - so no external JFET stage.
How the chip works internally is of cause not visible. It is even likely that the paths join before the variable gain part, as the variable gain can be quite some effort.
The coarse gain steps are externally, though only about a 1:7 voltage ratio for the 50 ohm  path. This does not direct match the gain steps for the user interface.
The 50 ohm input does not need much attenuation, as the maximum amplitude is limited anyway (possibly more than with other scopes).
For the 1 M path the later dividers have no matkings, the first seems to be 1:4  (750 K from the input side).

-snip
The front end chip definitely chooses between the different inputs. It looks like 2 x for 50 Ohm and 4 x for 1 M mode + an additional DC input that also has to be somehow combined with the AC part !
-snip

So the simple answer to my question is : Yes, we do not need a double path to ADC to provide 50 ohms input capability.

A agree with the yes, but not because the signals all go to the same chip.

The main point is that to combine the speed of both ADCs, one needs to send the same signal to both ADCs. So the 2 outputs must be equivalent and to both ADCs.
Similar the same PCB means the 4 unused links have to go to the unused ADC - so the 2 outputs are just to serve 2 ADCs, not to combine the 1M and 50 Ohm part later.
So enabling the 50 Ohm mode should be a SW thing only.

There may be a good reason in no enobeling it with the cheaper version:  the 50 ohm inputs are somewhat prone to overload.  With this scope this could mean damaging the front end chip and not just burning a 51 ohm resistor.

[markone]

-snip
There may be a good reason in no enobeling it with the cheaper version:  the 50 ohm inputs are somewhat prone to overload.  With this scope this could mean damaging the front end chip and not just burning a 51 ohm resistor.

I guess you mean 1K series is for mainstream "public" while 4K is for experienced, so 2K series (if it ever exists) would have single ADC chip but 50 ohms input and other 4K functionalities.

Changing focus, has any HDO4K/1K's owner changed the FANs ?

I noticed that they are controlled in speed, i wonder if it is under software control.

[Someone]

For all this debate about how the product might work....  has there ever been a "hack" on this forum that actually extended functionality by adding new code? As I recall it has always been enabling of already existing code by overriding/bypassing software lockouts.

[PA0PBZ]

I think there's been a few, one of the things that comes to mind is a 34401 which got additional functionality. I guess the thing is that manufacturers normally are getting the most out of the hardware, be it hidden behind additional licenses or a more expensive model, so if the code is already there it's easier to enable it than to write it yourself.

[bob808]

There not seem to have much choices. For HDO1K with A100T, only S2 bit2 can be used, as S4 need two ADC's, bit1 is for testing and bit3 is maybe for a future version with AWG.
Also in the frontend is a little difference that can be a filter to limit the bandwith of the frontend amplifier to 2-300 MHz. Tho I wish I'm wrong ...

Interesting that the PCBs are in fact not identical. I wonder if it's just a different revision for the same PCB or there's slight differences between the HDO1k and HDO4k models.

[Kleinstein]

For all this debate about how the product might work....  has there ever been a "hack" on this forum that actually extended functionality by adding new code? As I recall it has always been enabling of already existing code by overriding/bypassing software lockouts.

I faintly remember some modifications to the Solartron 7081 DMM to fix a problem. Not really adding functionality, but a late unofficial bug fix.  Still this comes pretty close.
For the R6581T DMM there is some effort to fix an often bad low level calibration: no real code change, but get acess to the low level CAL part to fix a common weakness.
AFAIK for the HP34401 one can get another smaller current range for AC, though a bit unclear with CAL constants. No added code but more enabling a never official option.

For the SDM3055/3045 there is something around for a more or less new firmware (not sure if this is for the inbound or outbound part or manybe both). AFAIK this came with additions and parts still missing.

Interesting that the PCBs are in fact not identical. I wonder if it's just a different revision for the same PCB or there's slight differences between the HDO1k and HDO4k models.

To me the difference looks rather minor, maybe a more normal PCB revison you may also find with the other version. The max speed for the 1 M input should not be that different and the change should mainly effect the highest V/div setting. So I don't think it has something to do with 1/2 ADCs, more like a minor flatness improvement.

[tautech]

For the SDM3055/3045 there is something around for a more or less new firmware (not sure if this is for the inbound or outbound part or manybe both). AFAIK this came with additions and parts still missing.

Not exactly.
There's a SDM3055 hack by a member to enable 6.5 digit measurements.
SDM3045X is missing from China where instead they have a 5.5 digit model called SDM3055X-E and it was my suspicion they were of the same HW but ran different FW versions. A few months ago this was confirmed and methods to flash back and forth between models developed.

[Martin72]

Changing focus, has any HDO4K/1K's owner changed the FANs ?

I noticed that they are controlled in speed, i wonder if it is under software control.

It´s possible to control fans which got only 2-wire connections, but kinda uncommon today.

[bob808]

if there's any GPIO's available you could probably build a small pcb to control the fans. 
temp information should be available in sys/class/thermal/ I think, so you could build a script for it. how many thermal zones are there?
edit: but you'd have to open it up, so you'd lose your warranty.

[markone]

if there's any GPIO's available you could probably build a small pcb to control the fans. 
temp information should be available in sys/class/thermal/ I think, so you could build a script for it. how many thermal zones are there?
edit: but you'd have to open it up, so you'd lose your warranty.

I was thinking at something different from making hardware modification, if the fan is actually software controlled with a buck converter, maybe there is somewhere in the file system a regulation table or regulation parameters set.

What i noticed is that when scope starts FANs run at low-medium speed while at the end of boot stage ramp-up to quite high speed, so quite high noise, in a room with 17-18 °C 

The aim would be to lower the darn FAN noise without opening the case (for warranty reason), otherwise those two jokes (original fans) would have gone day one replaced by some decent 60x25 devices (12-15 euros cad).

Could you give me some directions to sniff "sys/class/thermal/" ?

[bob808]

They might be voltage/current controlled then. Which would be good news.
get a shell on the scope and

Code: [Select]

ls /sys/class/thermalthere should be folders for each thermal zone. go in each and cat the "temp" file. hope it has this structure

[markone]

Changing focus, has any HDO4K/1K's owner changed the FANs ?

I noticed that they are controlled in speed, i wonder if it is under software control.

It´s possible to control fans which got only 2-wire connections, but kinda uncommon today.

This is a thing that I do in regular basis in dirty cheap projects, with massive productions a BOM's cost cut of few cent it is more and more frequent, so now you can find quite frequently double speed control with a dropping resistor and couple of transistors, where in the past was used a buck converter and 3 pins fan with speed reading, or even better a straight 4 pins DC FAN, all things that are becoming a luxury.

Of course you are "open loop" but if you respect the minimum FAN voltage (like 6-7V for 12VDC fans) you have not particular problem.

It's s***, i know, but it's the new normal for many commercial devices.

[bob808]

I looked at the photos for the backside of the PCB and to me those look like 2 SO8 MOSFETs and maybe a smaller sot23 one to drive their gates. It may be available in software. There may be clues in the dtb file.

[markone]

I looked at the photos for the backside of the PCB and to me those look like 2 SO8 MOSFETs and maybe a smaller sot23 one to drive their gates. It may be available in software. There may be clues in the dtb file.

What is the "dtb file" ?