40,000 Count Multimeter under $50 - HoldPeak HP-770D

[LooseJunkHater]

https://www.aliexpress.com/item/32983003087.html



This multimeter appears to have been out now for over 4 years now with essentially no attention on the EEVblog for hobbies. Per the linked video, it seems incredibly accurate for the cost (measuring down to 1uv) but lacks the safety features of a Fluke.

Anyone on the forum have one? I was thinking of buying a few more Aneng 8009's (~$25) but for double the cost, I get 4x the counts (9999 count vs 40,000 count) and additional features. Opinions for hobby use?

[coromonadalix]

everything is said in the video ....  and seeing the fuses .......  not


other thread

https://www.eevblog.com/forum/testgear/holdpeak-hp-770d-issue/

[Swake]

If 'hobby' means you're never going to exceed measurements of about 50 V then it is good enough. Above that it is not safe. This is valid for all cheap DMM's. I've got to use an Aneng AN9002 sometimes and, while it is actually very nice for the price, I will not approach it to anything like mains voltage because if I make just one wrong move it's potentially the last one I'll ever make.

Sounds likes this will be your first DMM. Don't give the specs and features too much importance. What you really need is a stable meter giving repeatable measurements. Repeatability is much more important than absolute accuracy or precision. Example: if you measure a value you want it to be in the ballpark of what you expect, that is close enough, you likely only need it to be within 2% or so. Any meter can do that. What you do care is that next time you measure the same value it effectively shows up as the same value so it gives you trust in what you measure and can be compared to another measurement.

Off course it is nice to have 0.1% accuracy, but you don't need that. Same for anything less than 10 mV, you don't care. And when you start caring about these things you need a much more elaborate lab.

[GigaJoe]

have one of this

"Incredibly accurate"  probably quite exaggerated

it has microvolts, but shorted it shows -7 microvolts now. usually it off from 2-3 LSN to 5-7 depending what you measure. limit 42000 useful to work with batteries where around 4V. (vs aneng 9999 )

To turn on backlit you need to stick an electrical tape on the top center , where the sensor is located.
speed is average \ mediocre - especially basic resistance algorithm switching down step by step ....

So overall it is an average thingi .

My daily usage is overclocked owon b41t+  (with 8M crystal instead 4m)  - it measures any resistance 0.4 sec delay - very fast , useful if you need to find one in a haystack.  it 22000 but precise to  1 LSD ;   it a bit off more as i add more tvs diodes, thermistors , varistors. GTD.
 
 
 

[magic]

This multimeter appears to have been out now for over 4 years now with essentially no attention on the EEVblog for hobbies. Per the linked video, it seems incredibly accurate for the cost (measuring down to 1uv) but lacks the safety features of a Fluke.

Can't be assed to watch the video.
Does it actually measure voltages like 1μV or -3μV or does it round everything less than ±5μV to zero in order to hide its offset voltage?

This meter has been reviewed at lygte-info.dk and I remember finding something about it that made me prefer the Aneng AN870 as a cheap, high-ish resolution handheld.

[iMo]

I saw the video and indeed it has got the 1uV resolution, no rounding around +/-5uV (stepping from 0 step 1uV shown)..
Usable like null meter or something like that, imho..
BUT, the 1.2V reference there is the LM385-1.2, afaik, and the input dividers are discrete resistors..

[magic]

I saw the video and indeed it has got the 1uV resolution, no rounding around +/-5uV (stepping from 0 step 1uV shown)..

OK, so that's actually better than the Aneng rubbish. Even if there is offset, you can cancel it with the "zero" button, but you can't cancel cheating by the firmware

However, its millivolt range is low impedance - I think this is what I didn't like about it. I wonder if it could be modded perhaps?

BUT, the 1.2V reference there is the LM385-1.2, afaik, and the input dividers are discrete resistors..

That's standard in similar meters.
Resistors are not too bad and the meter has built-in resistor calibration function called the "ohms range"
The voltage reference could possibly be upgraded and the meter recalibrated if you are really nuts.
IIRC most of these meters have a cal jumper and usage of the cal mode is known and understood.

edit
On second, I'm not sure if you can use the ohms range to calibrate internal resistors against an external standard.
The chip may have measurement errors, or may apply corrections programmed during factory calibration.

[iMo]

at 6:26 starts the playing with the 1uV step around zero..
I think a nice meter for use on the bench when messing with low-voltage low-power electronics (like max 24V and max 1A). It has got 10nA resolution, 1uV resolution, 10mOhm resolution (on the lowest ranges, sure) as I can see in the video. Would be great to know how to calibrate it, there is a standard 8pin flash storing the data..
A pity they use the 9V block battery when the Vreg is 3.3V one inside, it takes 5.6mA (backlit leds off) what is not particularly low for a, say, 9V 250mAh NiMH battery, imho..
At least a step-down switcher would be handy there with the 9V battery.

[LooseJunkHater]

A pity they use the 9V block battery when the Vreg is 3.3V one inside, it takes 5.6mA (backlit leds off) what is not particularly low for a, say, 9V 250mAh NiMH battery, imho..
At least a step-down switcher would be handy there with the 9V battery.

I actually think it's kinda nice that it uses a 9v battery with a 3.3v regulator. It allows the meter to be easily modded to use rechargeable lithium batteries in a 2s configuration. It'll have a longer battery life compared to the 9v cell and be rechargeable.

[LooseJunkHater]

My daily usage is overclocked owon b41t+  (with 8M crystal instead 4m)  - it measures any resistance 0.4 sec delay - very fast , useful if you need to find one in a haystack.  it 22000 but precise to  1 LSD ;   it a bit off more as i add more tvs diodes, thermistors , varistors. GTD.

Never heard of overclocking a meter before. I can see that the HoldPeak also has an internal crystal, so I wonder if it also can be overclocked? Aside from increased power consumption, will the overclock significantly impact the lifespan of the meter or effect measurement accuracy? There doesn't appear to be a lot of information on this; nothing on YouTube, and only one small thread on this forum.

https://www.eevblog.com/forum/testgear/multimeter-overclocking/

[coromonadalix]

mmmmm  the HY3131  is a 50 k count chip,   maybe there could be some hack ??  eepom edit     someting

[GigaJoe]

Never heard of overclocking a meter before.

basic idea to change quartz , of course all related  measuring changing as well, additionally stability on lower power voltage  , can hiccup.  some aneng works on 3x ( 12M) quartz .  never play on expensive meters , due to mostly , complexity of cirquit, and no reason, if you already have cheap overclocked one.

basic modding to add polymer caps with battery usually very noisy power line  generated by chip itself, caps to vref , soldered connectors,  something else like GDT

[iMo]

A pity they use the 9V block battery when the Vreg is 3.3V one inside, it takes 5.6mA (backlit leds off) what is not particularly low for a, say, 9V 250mAh NiMH battery, imho..
At least a step-down switcher would be handy there with the 9V battery.

I actually think it's kinda nice that it uses a 9v battery with a 3.3v regulator. It allows the meter to be easily modded to use rechargeable lithium batteries in a 2s configuration. It'll have a longer battery life compared to the 9v cell and be rechargeable.

Something like this..

[iMo]

mmmmm  the HY3131  is a 50 k count chip,   maybe there could be some hack ??  eepom edit     someting

[LooseJunkHater]

basic modding to add polymer caps with battery usually very noisy power line  generated by chip itself, caps to vref , soldered connectors,  something else like GDT

Not sure what "GDT" is, but essentially just add some ceramic caps to the output of the linear regulator, monitor for heat (using a thermal camera), and see if it works? Can I essentially use any crystal which is faster than the original?

[LooseJunkHater]

Something like this..

Exactly, except I'd personally do 2s (one lithium cell in each aa battery slot) due to the device using a (per video) 3.5v linear regulator.

[magic]

A pity they use the 9V block battery when the Vreg is 3.3V one inside, it takes 5.6mA (backlit leds off) what is not particularly low for a, say, 9V 250mAh NiMH battery, imho..

Now, this was probably the main reason I disliked it.
Very wasteful, lousy battery life.

I actually think it's kinda nice that it uses a 9v battery with a 3.3v regulator. It allows the meter to be easily modded to use rechargeable lithium batteries in a 2s configuration. It'll have a longer battery life compared to the 9v cell and be rechargeable.

I would even consider single cell. The regulator is probably LDO and IIRC Li-ion delivers almost all its charge at more than 3.3V.

The "battery low" indicator would need to be modded either way, because it reportedly trips at 8V.

[LooseJunkHater]

The "battery low" indicator would need to be modded either way, because it reportedly trips at 8V.

I'd imagine it's just a simple voltage divider. Considering how few components exist on the PCB, I couldn't imagine this will be problematic to locate and modify.

[GigaJoe]

Can I essentially use any crystal which is faster than the original?

certainly,
some dmm has 2 chips and 2 crystals - don't recommend touch those , most likely fail

[LooseJunkHater]

Can I essentially use any crystal which is faster than the original?

certainly,
some dmm has 2 chips and 2 crystals - don't recommend touch those , most likely fail

https://www.hycontek.com/wp-content/uploads/DS-HY3131_EN.pdf

Page 20 of the schematic does make mention of an internal crystal oscillator. I wonder if the internal crystal oscillator has been disabled for this product so that an external one is used instead?

[iMo]

That means "an internal oscillator for an external crystal"..
There is not such thing like "internal crystal osc" in that chip..

[LooseJunkHater]

Just ordered the meter for $44 USD. I'm expecting the meter to be pretty slow and if it is I'll attempt to overclock it.

[iMo]

I would even consider single cell. The regulator is probably LDO and IIRC Li-ion delivers almost all its charge at more than 3.3V.

The "battery low" indicator would need to be modded either way, because it reportedly trips at 8V.

There is a diode in series with the + lead of the battery (see the pcb above), so it will not work with a single cell (unless you remove the diode). Btw. there are lipo 9V rechargeable with a lipo cell inside together with charger and switcher (step-up). Perhaps you can hack the switcher to produce 4.8V for example.. 

[siealex]

These switchers produce enough noise to screw up the whole device.

[geb]

I've been fiddling with my HP-770D for months.

If you're looking for great accuracy, even with calibration, you'll be disappointed. At least I am, with my one sample of this DMM.

It initially drifted a lot. I replaced the LM385 with an ICL8069A and it got worse. Replaced the LM385, cleaned the PCB, removed any paralleled capacitors recommended by other forums. It's better now, drift-wise, but not perfect.

Someone in a forum (here? Kazus?) recommended a 55000 count mod. I tried it. The linearity isn't great across any given scale, but with the increase from 40000 to 55000 it magnified the nonlinearity. There's a few (perhaps 4?) digits of inaccuracy as I go from 1V to 4V, but the inaccuracy increases to 21 digits at 5V. As in, depending on whether I calibrate at 1V or 2V or 3V or 4V, measuring 5V afterward is anywhere from 5.0006 to 5.0021. I apologize if I'm not perfectly clear on this one. I didn't keep any notes.

It calibrates by using basically the same method as ZT-102, AN8002, and other cheap meters using a Hycon MCU. Close the jumper, turn on the unit and switch its dial quickly to ohms. Wait for its self-calibration to finish; it'll beep in mid-calibration and again at the end. Use HOLD and RANGE to decrease and increase, and SEL to advance. One word of warning: If you are displaying 40000 and you increase, the calibration routine will fail and you'll need to reload the EEPROM from a backup.

Making a backup of the EEPROM: it's the same process as other Hycon MCU meters. You must move the dial away from the OFF position. You'll cause the MCU to run, and it'll interfere with your backup. Short out the outer two pins of the six-pin unpopulated inline connector near the calibration jumper, which will hold the MCU in reset so you can read the EEPROM.

Like GigaJoe, my daily grab-and-go meter is an Owon B14T+ with the DCV 10-turn pot painstakingly calibrated. I didn't overclock -- I like how fast its autoranging is already. (It has a little nonlinearity across 2.2V-22V but I'm living with it. ) I'd overclock my original ZT-102 if I knew how to change the divider on its Hz mode. I can't recall seeing an option for that in the Hycon EEPROM spec.