HMP4040 calibration and adjustment

[PolarisAT]

I have a HMP4040 sitting on my bench and it's a little bit off from its spec. I'm basically very happy with it because it's astonishingly precise and i got used to it being spot on. But now it's off and I would like to adjust it back to being perfect again. The manufacturer does not provide any GPIB commands or tools to do this. Does anyone of you know how to do a proper calibration and adjustment on these devices?

Thanks!

[branadic]

The only hint to some command can be found in the manual.

-branadic-

[PolarisAT]

Hi branadic,

what you are referring to is the status bit of the calibration. I can poll it and it it set to 0 as expected. The manual says that these bits get set if the state of the related bit is "questionable".
I don't expect my calibration to be "broken". I just want to set the calibration factors to new values but I can't find any service manual containing the related commands or a proprietary tool to do the calibration with.

I attached some pictures of a quick teardown if anybody is interested.
To get into the device you have to unscrew 6 TORX10 screws on the back. One of them is hidden behind the "warranty void" sticker and another one hides behind the communication port.
Additionally you have to remove the two fand covers on the sides of the case. Then you can push the inner part forward to get it out of the housing.

The PCBs look nice, but sparingly labeled. Components are of good quality and known brands (Nichicon caps, EPCOS chokes) and a lot of ATMEL ATmega processors.

[wildbill]

Hi guys,

i have >30 HMP4040 in my lab and every year we perform the calibration.
Actually we didn't found any of them out of spec in any of the measurement.

When a well known calibration company used a "wrong" calibration procedure, many units were declared out of spec
and the adjustment was "required". (when we found that the procedure was wrong, R&S declared them in spec).
They were not able to adjust them, they checked from Rhode&Schwarz (that acquired Hameg) the tool availability and is not sold to third party companies.
Adjustment is possible only at R&S.
We checked also with another calibration company if the tool was available and they confirmed it is not.

[PolarisAT]

Thanks wildbill for your insight. I/we also operate >80 HMP series supplies therefore we would have been very interested in doing the services inhouse.
But if the required tools and information is not given to 3rd parties I can stop looking for a solution now.
But all in all the HMP40x0 series is a very nice tool and as I mentioned also very precise.
We only had 3 mishaps so far. One went up in flames over night 
And on the other 2 one of the output channels broke. But that was most likely our fault.

[electr_peter]

HMP4030/HMP4040 firmware update procedure updates "front panel" and 3/4 channels separately one by one. Communication via USB/serial is only to the "front panel" board while calibration coefficients are likely in each channel's MCU. Without special R&S tools it is not feasible to calibrate (adjust) the unit.

[electr_peter]

We only had 3 mishaps so far. One went up in flames over night

I would like to hear more details on this. Did PSU fail while under load or just turned on? And did it generate flames?

[PolarisAT]

It was turned on and under little load, maybe a few watts. It did burst into flames and luckily another device was placed on top and it helped contain the flames shooting out through the top vents. It did not trigger the fire alarm, probably it didn't smoke very much. It was during the night, so nobody knows what really happened. Only the burning marks told the story. Both devices were a lost case but nothing else in the room was affected. HAMEG was very helpful back then and replaced the device immediately.
Since this day our company has a rule to not leave a HMP running over night unattended if possible.

[YetAnotherTechie]

With >80 power supplies and they don't provide info on calibration? that's rough, i can understand it for very high tech equipment, but for a power supply feels wrong. I had a look in the software update and the firmware has the scpi commands *SETCALMODE *CLEARCALMODE, you could try it on the one that caught fire  but without the rest of the procedure it will likely yield bad results.

[PolarisAT]

I had a look in the software update and the firmware has the scpi commands *SETCALMODE *CLEARCALMODE

Nice! Where did you find it exactly? Was it in the FW update tool or the firmware itself? Is there anything other that looks useful?

[F64098]

We only had 3 mishaps so far. One went up in flames over night

I would like to hear more details on this. Did PSU fail while under load or just turned on? And did it generate flames?

I lost a HMP2030 two years ago, when running it constantly for a full week with minimum load. The PCB (4 layers) beneath the rectifier blew up due to an internal short, developed over the time.
I got no help from R&S/Hameg and can't buy spare parts for a reasonable price.
It was replaced by a brand new HMP4030 (sold for 50% by Conrad Electronics) but there's still the apprehension of growing the next luxurios paperweight 

But yes, it's a really cool (or hot  ) unit. I really like it!

Best regards

Frank

[YetAnotherTechie]

I had a look in the software update and the firmware has the scpi commands *SETCALMODE *CLEARCALMODE

Nice! Where did you find it exactly? Was it in the FW update tool or the firmware itself? Is there anything other that looks useful?

Not really, but i just glanced.
Inside RS_HMPUpdate_FWV2.70_5800.4292.00.zip is a file called FW1A.hmp that has those commands.

[PolarisAT]

I checked the file myself and as you mentioned the commands are listed. Interestingly they are listed together with all the other SCPI commands that are supported. Even the answer to IDN? is listed. For me it looks like this file is a combination of some constants, maybe some config stuff and the AVR firmware. I ran the file through an AVR disassembler. It does give some meaningful output but not much. Either the code is very short or the disassembler would need some more manual help to cope with the file format.

Because there are 5 firmware files in the bundle one might get to the conclusion that 4 of them are for the individual channels and one is for the front panel. But then you would expect 4 of these files to be identical or at least very similar. But they are not as far as I can tell. Only one of them sticks out because it contains the SCPI command list and some other constants that are familiar. The update tool is written in C++ as far as i can tell, decompiling it might give a cloue on how the files are used but I'm not a c++ or decompiler expert.

If I have some time left I will try to sniff the communication during the FW update, it might give some insight in whats going on.

[hameghm80xx]

Hi there,
I am a Hameg collector and I mainly collect the old HM80XX series. In this series there are some devices of the last generation (Hm8012, Hm8015, Hm8018) in which the calibration routine is initiated with a combination of buttons when switching on (hold down two buttons at the same time when starting the device).
With the HMP4040, the firmware update is also initiated in the same way. Maybe there is such a routine here too?
Has anyone experimented with it yet?
I have a HMP4040 too and there is also a smal (but anoying) error in the mesurements.

With the HM8012 you then have to select a measuring range, apply a specified reference value (e.g. 5,000V) to the terminals and then save this reference value as the new reference value for this measuring range by pressing a button.

Marc