Hameg HM8042 Curve Tracer Any users out there?

[nevetslab]

I must have bought this Hameg HM8042 Semiconductor Curve Tracer in a dual-mainframe HM8001 back in March of 2009. Added a pair of BNC connectors on the rear panel of the mainframe so I could drive the Tek 7633 scope in XY for proper display. Over the years, I've used it for screening Power MosFET's used in the Ampeg SVTx-Pro series amps along with other small signal devices when I needed matching or specific range of the xstr or FET characteristics. I had always had issues with the Rotary Encoder in setting the instrument up. And, it was always a real PITA to change from NPN to PNP or N-Ch to P-Ch devices.

Having just swapped out Tek Scopes, now having a rack-mount Tek 7623A in place of the 7633 (still lacking CRT display), I had to re-label the X & Y BNC cables from the mainframe. That's when I found the Hameg Encoder is now really not working right. I did some IC reseating, and made the mistake of NOT properly grounding myself with my wrist-band grounding straps. When I fired up the Hameg again, it's now got problems.

I found I had the schematics of the instrument at the end of the multi-language manual, but, I've also found the date of those schematics being May 10, 1994, while the date codes of most IC's range from 9930 to 9940, and the MFGR date seems to be early 2000. There are a number of IC's that are no longer used, and others in their place on this instrument.

The Encoder, it being a 3-terminal part, no Switch, Horizontal mount off of the Front Panel PCB....it's pin-out appears NOT to be the same as most common inexpensive 3-terminal Encoders. The common pattern is A-C-B, with the middle terminal 'C' being Ground, and A & B being the two 'phases' of the encoder output. This Encoder that's installed on the Front Panel PCB has the 'A' terminal grounded, and the two phases are off of the 'C and B terminals. Or, perhaps this is a different form factor for an Encoder, which I have yet to come across while searching for that form factor. I don't yet know whose Encoder they selected. So far, I have NOT been able to get the molded plastic Front Panel frame free from the PCB, which would be required to replace the Encoder.

I've gone and ordered new IC's that are in use on this PCB assy of the Curve Tracer. Hopefully I'll be able to restore this to working status, as it presently is not working...which all started with the continual decline of the Encoder NOT behaving properly.

My iPhone 5 'camera' is presently over at an iPhone repair shop getting a new battery installed, so I wasn't able to wake it up this past week to get detailed photos of the insides. I do have these images from the internet as well as previous shots in use during screening MosFET's.







It's limited to a maximum of 40V Vc or Vd, so it won't get you out to breakdown voltages of most high voltage devices, but, it has been handy once I got the instrument set up, and had been yielding very repeatable results. I never had the funds for a Tek 576, which can also be set up to work with vacuum tubes. Getting that front panel assembly separated from the chassis, so I can remove/replace the Encoder is one of my goals....though it could just be the MC14490P CMOS DeBounce IC that follows the Encoder output. Still puzzled with the 'A' terminal of the Encoder being grounded, while the middle C-terminal is wired in the layout as one of the two terminals that feed the MC14490P.

Schematic is at the back of this Owner's Manual.

HM8042_D_E.pdf

[nevetslab]

Today after disconnecting the ribbon cable from the 11mm 15PPR Rotary Encoder from the circuitry, I hooked up my Fluke 8060A 200 ohms range, zero'd it to watch DCR and added the beeper and looked at the DCR readings connecting between the terminals. The normal configuration we see in today's 11mm Encoders like this, facing the lead breakout...left terminal is 'A', middle terminal is 'C' and right terminal is 'B'. Normally wired with 'C' terminal grounded. In this Hameg HM8042 product, the 'A' terminal is grounded and the 'C' and 'B' terminal are fed to the MC14490P Debounce IC.

I watched the DCR readings as I rotated the shaft one step at a time. 30 detent, and reads as a 15PPR (15 pulses per rotation). Looked at the readings connecting 'A' & 'C', 'A' & 'B', and 'C' & 'B'. Seemed like the readings were more consistent using the 'C' terminal for GND than using the 'A' terminal as the PCB is laid out. But, I did get readings less than an 1 ohm at each alternate step. Many positions I had to rock the shaft back and forth to get reliable readings.....which relates to the inconsistent results I've been having with that encoder all along. That's what I was afraid of.

I found three products that look like a suitable replacement, though what's available in distribution in 30 detent 15PPR with 6mm metal shaft...all are D-shaft, and the overall shaft length is a bit shorter than the solid split 6mm shaft used in this product. I bought a Wurth Elektronik WS-ENTV encoder and a Bourns PEC11L-4220F-S0015 Encoder. The main difference between the two is the Bourns has a 100,000 cycle life instead of the 15,000 cycle of the Wurth (as well as an Alps EC11E series encoder). I should have the two new Encoder's by the end of the week. Getting that front panel Plastic Molded part to separate from the chassis looks to be a real PITA. This package is designed to be mfgr'd and assembled....NOT to be disassembled.

Hameg was purchased by Rhode & Schwarz in early 2000, and products thru Hameg and R & S were discontinued in 2011, I think it was. I haven't tried contacting them to see if a later updated schematic is available. Interesting that in the schematic I posted, the Encoder, which is found on the Front Panel Display Board schematic, in the Connector I/O portion of the drawing. The Encoder is called out as DG401, P/N MPS20. Its' configuration is the common 'A-C-B' configuration that is commonly sold in distribution, but is NOT wired that way on the PCB. DOES THAT MAKE A DIFFERENCE???

Bourns_PEC11L Series Encoder.pdf

[Helmholtz]

https://www.rohde-schwarz.com/home_4...?change_c=true

[nevetslab]

Today, the Wurth Elektronik WS-ENTV Encoder came in, so I was able to see how this part behaved. First I checked the DCR readings between the 'C' (middle) terminal and the 'A' and 'B' terminals as I slowly rotated it thru a couple cycles. 15PPR. After setting the Fluke 8060A DMM for 200ohms, zero'd it for REL measure, having 10 mOhm resolution, I was getting consistent <60 mOhm readings at each step (typcially reading 50 mOhms). When I then used the 'A' terminal as Common, and measured both the 'C' and 'B' terminal, I got slightly higher readings...<70mOhms reading 'A' to 'C' and 'A' to 'B'. This is lower than I was getting on the installed Encoder of the instrument that has been eratic as long as I've owned it. So, I definitely need to replace the Hameg Encoder based on how a new part reads.

I then made a tiny test fixture using a cheap IC socket, that could accept the three terminals of the Encoder, socket having mating surfaces on both sides of the Encoder terminals. I soldered in three 12k 1/4W resistors across the pins (1 to 14, 2 to 13, 3 to 12), which then allowed me to connect two scope probes to the Encoder terminals and fed it with +5V common to both the 'A' and 'B' terminal to begin with, and GND to 'C', which was also where I connected my scope probe GND to.

Put the scope in storage mode, had to set it for single triggered sweep (Line Sync), set the time base for 100mS, and set the Auto-Erase to save the single-sweep to display a number of pulses and hold it. With 100mS/Div sweep, I could see the Quadrature offset of the 'A' and 'B' pulses both in CW and CCW rotation. The Pulse length was equal, between 'A' and 'B', and 'A' was leading B' in turning on and off.

When I then set the 'A' terminal for GND, and fed my 5V to pins 'C' and 'B', in CW rotation, 'C' lead B', turning on, while BOTH were turning off at the same time. In CCW rotation, 'B' & 'C', turn on at the same time, while 'B' turns off first, AND the pulse length is shorter on 'B' than on 'C'.

I still am without a camera to capture the scope images, but should have that back in a few days. So, as I suspected, the pulse waveforms are NOT the same using the 'A' terminal for common GND than using the middle terminal 'C' for GND as is called out on the products we can purchase from different mfgrs.

I won't know what the Encoder Hameg has installed until the Bourns Encoder arrives, and I again attempt to remove the front panel frame/Escutcheon assy from the chassis. It looks like I will have to unsolder the three Banana Sockets for the Emitter, Base and Collector terminals. I don't yet know if I have to also unsolder those tiny front panel 'X' and 'Y' terminals as well. Hope not, as those are just NOT accessible.

Most of the replacement IC's came in, while I'm still waiting for the last few to come in from China and the UK.

More to follow...........

[nevetslab]

https://www.rohde-schwarz.com/home_4...?change_c=true

I checked their website, but wasn't able to find the HM8042 in their Hameg listings. I haven't yet contacted their service dept to see if I can obtain later generation schematics that reflect the parts I see in my instrument, which is different than what's on the May 10, 1994 schematics, which is all I find on the internet.

[nevetslab]

Today I got this Curve Tracer completely pulled apart. I did manage to break off the 4-wire AC cable to the Xfmr PCB assy, but did get that repaired. After unsoldering the three Banana Jacks from the rear side of the FP PCB assy, and unsoldering the tiny 2-wire cable that carries the X-Y Signal over to the main PCB, I was able to carefully pry the front panel assy with it's molded plastic part &amp; FP Escutcheon away from the FP PCB.








The key to separating the FP PCB/Panel assy from the chassis & Main PCB is unsoldering the three Banana Jacks 'E/S', 'B/G' &amp; 'C/D'. Had to first unsolder the RED, BLK and shielded cables from those connectors. And unsolder the tiny GRN loop-connector from the X-Y connectors of which I've never seen a mating connector. Scope probes don't fit them, but will yield signal. You just can't place two probes that close together.




I carefully unsoldered the Encoder from this front panel PCB assy, and found it was an Alps Encoder, though no P/N or date code on it. I did measure it's DCR step by step...it being a 30 detent 15PPR part, and a 25mm long Encoder shaft, which is a Factory Special...can't be purchased thru Distribution. Now when I had tested the Encoder while it was still soldered into the PCB, while having disconnected the 10-cond ribbon cable parked just above the Encoder to the MC14490P Debounce IC, I was reading very inconsistent ON readings. With this part removed, I was getting consistent ~50mOhm DCR readings in the ON-settings. Tomorrow morning, I'll set up my little test fixture to spin the shaft and catch the Pulse train in both CW and CCW directions using the middle 'C' terminal against the 'A' and 'B' Terminals. The PCB is, however laid out for GROUNDING the 'A' terminal, and routes the 'C' and 'B' terminals to the system. I'll still replace this control with a fresh Bourns part that has 100k rotation life. The Alps and others I had looked at only had 15k rotation life.



This post crashed while I was posting the last two images, so I was able to recover it, but have odd little symbols to the right of each photo. Hopefully it will post ok.

I think I have installed the NAIS TQ2-5V DPDT relay upside down. I'll verify that in the morning when I get back to the shop. The 'BAR' indicator is where the relay coils is Pins 1 & 10, and the opposite end have NO Connections Pins 5 & 6. This relay is NOT shown in the manual's schematic in a 10-pin configuration. NC contacts would be Pins 2 & 3 and Pins 8 & 9. I see connections of the Base and Collector going to Normally Open connections, and the solder pads going to the middle connections, which should be the Poles of the relay. The Emitter Banana connector is Grounded. So the relay is engaged to connect the Base and Collector to the test circuit, if I'm reading this right.

My shop computer has now crashed, and I can't get to the internet. How bloody thrilling!

[nevetslab]

Today I checked out the worn Alps Encoder I had removed from the Front Panel PCB of this Curve Tracer along with that of the new Bourns PEC11L-4220F-S0015 15PPR Encoders by applying +5VDC to the two terminals of the Encoder with the normal middle terminal grounded, and then that of how Hameg laid out the PCB to instead ground the 'A' terminal of the Encoder and feed the 'C' and 'B' terminals to the Curve Tracer system. While measuring the used Alps Encoder with my Fluke for measuring the DCR, what I saw turning the control and catching the pulse train on the storage scope revealed the defects of the worn Encoder.





These captured pulses were better then most of the test shots I saw while trying to get consistent output pulses. It still shows Quadrature (A leads B or C leads B when the 'A' terminal is grounded as is laid out in the front panel PCB. All Encoders you look at in mfgr's doc's show the middle 'C' terminal being Grounded.

I then replaced the used Alps part with the new Bourns part:





The test 'fixture' was a simple 14-pin IC Socket with three 12k 1/4W resistors soldered from one side to the other side so I could tie the two resistors used to the Encoder Terminals, while grounding the other terminal, whether it was the 'normal' wiring as data sheets call out, or wired how Hameg installed those 'standard' Encoders.

I wasn't seeing any eratic pulses like I was getting on the aged Alps Encoder I had removed.

I got the front panel assembly apart and installed one of the new Bourns Encoders into it, finding there was still 4.5mm of the D-Shaft protruding past the front panel Escutcheon, so standard 15PPR/30-detent Encoders do work for a replacement part.



I'm still waiting for the rest of the replacement CMOS IC's to arrive. I did reassemble the front panel, did find I had indeed put the relay in backwards when I had removed it to suck out solder from one of the wire holes.

Next thread shows the reassembled unit, less replacing IC's.

[nevetslab]

Reassembly of the Front Panel and chassis with the new Bourns Digital Encoder:





There was a little bit of a skirt left on the back side of the Encoder Collet knob, so I filed that down flush to the mechanics of the collet assy to get a touch more of the Single-D 6mm shaft into that knob. Hopefully all this work with swapping out some of the CMOS IC's will get this unit back up and working right.

[amsoilsynth@msn.com]

We’re you able to save this 8042?

[amsoilsynth@msn.com]

I’m having to take care of the same encoder on my Hameg HM6042 (HM8042 with a display) along with a few other issues.

[amsoilsynth@msn.com]

Minor issues though.

[nevetslab]

We’re you able to save this 8042?

I picked up the replacement parts for it, and have replaced the defective encoder. I've been busy with my daily-dirt shop work lately, and haven't stopped to swap out suspect IC's in the unit. When I get back to it, I'll post the results. So, you have the CRT version of this?

[nevetslab]

We’re you able to save this 8042?

Last was working on this 11/22/2022, or at least that was the last entry in my Service Notes on the project. I had picked up most of the replacement IC's that I found were installed on the version of the Instrument I have, which I don't think quite matches the schematics I have in the owners/service manual. I had made what appears to be a major mistake when I was trying to get the unit back up and running on 11/02/2022: That being I didn't ground myself with a wrist band to Bench Ground, nor did I place an insulated ground sheet underneath the instrument while one IC at a time, was lifting each part up and out of it's PCB socket and then reseating each one. Did all IC's thruout the instrument that I could get at. I hadn't yet assessed just how to remove the rear panel mainframe Connector PCB until this past week.

When I had applied power in this period after having successfully replaced the Digital Step Pot, I was getting major errors in having the instrument function as a curve tracer. I set it aside until June 13, 2024 last week. I began swapping out IC's carefully with replacement parts I had purchased in November 2022. I found I hadn't ordered enough of the LF412 Low-drift JFET Dual Op Amps, and just ordered more this past week.

After I got everything swapped out that I could, I powered up the instrument, which, not having a plug-in extender cable fabricated, I can't probe inside the instrument without removing the Hameg Mainframe from the stack of gear that is sitting on top of it. Major bench disassembly required to do that, so for the time being, having to wing it without being able to probe inside.

What I've found is NOT encouraging. My baseline is no longer a flat line extending out to 40VDC (when Voltage is set for 40V). It's a slowly rising curve out to 20V and the first minor graticle line above baseline. then it rises from 20VDC to 40VDC to 1.2 divisions above the baseline. I DO NOT KNOW WHAT IS CAUSING THIS, AND NEED TO FIND A SOLUTION TO THIS.



i had an MPS A43 NPN TO-92 Xstr installed in the test cable for the curves, and on the baseline curve, Switched off DUT. So, THAT IS, AT PRESENT, the DC Baseline, which is, of course NOT CORRECT. So, I've only begun the digging into what problems I may have caused a couple years ago when I went thru the process if removing and re-installing the IC's from their sockets, using my two Optima 16-pin IC Removal tool along with a revised one for 8-pin IC's. I have NOT replaced any of the NPN, PNP or FET's in the instrument, apart from the power supply parts.





I'm presently in need of this instrument to screen the newly-acquired Exicon ECX10N20 N-Ch Lateral MosFET's and ECX10P20 P-Ch Lateral MosFET's for a Mesa Pulse 600 Bass Amp which is full of BUZ901P's and BUZ906P Lateral MosFET's.

I may have to work with this known baseline error and see what I can come up with.

At least the Digital Rotary Encoder is working properly, so I DID achieve that much, while having inadvertenly caused this baseline shift problem. Sure could use some eyes and intelligence in getting further into this problem to restore order to how it USED TO DISPLAY (flat baseline and curves reflecting what the Semi's under test are doing!).

I've attached the Manual for reference.

HAMEG HM8042 SEMICONDUCTOR CURVE TRACER SERVICE.doc

[nevetslab]

Yesterday, I looked thru the parts I had replaced earlier this month (having purchased them back in Nov 2022). Also purchased the rest of the LF412CN parts, just in case. After replacing the original LF412CN parts with the new ones, I got this E3 error code, so went back one IC at a time that I had swapped out and put the originals back into place until the error code was gone. But stil had the baseline error. Then looking at the schematic, in the area of the Ic/d Measurement, I found the original AD633 Analog Multiplier and plugged that back in, removing the replacement part I had installed. That appears to have corrected the baseline error/curves that has left me with uncertainty. Not having the calibration procedure for this instrument, it's been touch and go since having replaced the rotary digital encoder back in Nov 2022.



The Semiconductor plugged in is an MPS A43 NPN. So, it looks like I have a functional Curve Tracer again, it's display curves looking like they did when I was last using the instrument for screening and matching IRFP 240 & IRFP9240 N-CH and P-CH MosFET's. I'll follow up with those results, now needing to look at and select the newly-acquired Exicon ECX10N20 N-Ch Lateral MosFET's and ECX10P20 P-Ch Lateral MosFET's for a Mesa Pulse 600 Bass Amp which is full of BUZ901P's and BUZ906P Lateral MosFET's.

Onward........

[nevetslab]

This afternoon, finally having time to tackle the Mesa Pulse 600 Bass Amp that had multiple BUZ901P and BUZ906P N and P-Ch MosFET's fail on it, along with blown up 0.20 ohm/2W Source resistors, I pulled that heat sink assembly apart, removed the #6 mounting hardware holding the MosFET's to the inside surface of the heat sink assembly and then unsoldered all the MosFET's and source resistors. I went to test the one of the other replacement MosFET's I had picked up on the Hameg 8042, only to find I can't set the Gate voltage as needed, and the reading of that voltage on the instrument is NOT what I'm measuring on the Gate-Source terminals of the test leads connected to a suitable MosFET. RATS!

So, the task of fully restoring this Hameg HM8042 isn't yet complete after all. I had hoped that task was completed.

Guess having picked up one of those Peak Atlas DCA75 semiconductor test instruments wasn't such a bad decision after all. I did use it to find five of the BUZ906 P-Ch and 2 of the BUZ901P N-Ch parts survived, the rest shorted.

So, finishing up installing the replacement Exicon MosFET parts in the morning and hopefully getting that amp back up and running.

Any owners/users of the Hameg HM8042's out there who know their way around the insides?