I wish I could send my 184 to you for a fix. I've been working on it off and on for 4 years and can't figure out what's wrong. There's no signal output at the 50Ω out and tracing backwards on the board or the schematic I can't even figure out where it's supposed to originate. The only thing wrong I've ever found was the opamps in the power supply circuit were bad, so replacing them brought back the ±15V. Every transistor tests good, every cap tests good, there are no shorted resistors (the 3.9MΩ R14 in the VCG amplifier tests at 4.4MΩ but I don't think that would cause the problem), and every IC I've been able to test by swapping with a new one seems good. Any ideas?
Take a look at service manual, there is a block diageram before section 4. and the circuit diagrams in section 7 and board layouts too) are broken down into the mmatching logical blocks which helps isolate various parts for testing. Everything starts with positive and negative current sources (which are set by the VCG amp) and a current switch alternates between them based on timing and the hysteresis circuit, now constant current into a timing capacitor gives a linear slope and since it alternates +/- it gives a triangle, the speed of the rise/fall depending on the amount of current. This triangle goes to a buffer amplifier (from which we get a triangle) and this then feeds back to the hysteresis switch, which in turn triggers the swapping of the currents again so the slope goes the other way. *The TTL pulse output is taken from the Hysteresis section of the circuit so misses out the pre-amp and the output amplifier - so this is a handy place to check if you have a signal that is not getting amplified, or no oscillation at all. * The triangle from the buffer amp goes to the waveform switch (for triangle) and the sine converter witch sends a sine wave to the switch) The square wave and pulse are connected to the waveform switch come from a square wave shaper which is in turn from the hysteresis switch, but the switch has a DC position so that handily bypasses almost everything and lets you check the output amplifier in isolation! In terms of further fault finding it is all about eliminating whole sections, but first make certain you have all 5 voltages present - you say you ahve +-15 but check this is getting to the op amps e.g. IC 11 and 12 1. As well as +-15V check the +5, +10 and -5V supplies. The 5V uses a 7805 Regulator and can be checked on the logic ICs. The +10V, -5V are generated using a divider network and appear on the emitters of Q22 and Q23 so there are just 3 resistors and 2 transistors if those voltages are not present. you should also sheck the voltages reach IC6 ins 7 and 4, it might be a round can, if the board is positioned so that the 15V power supply is top right then the two transistors and the IC6 are at the bottom right corner. 2. Make sure the switch tops are not rotated on sharft and choose a constant signal that does not require any external trigger - easy fault to miss! 3. Check the pulse out - if it is present then the fault must be in the pre-amp or later. If it is present then try a DC waveform and vary the amplitude and see if that appears on output (no pulse = something before preamp, pulses and DC voltages = preamp, Pulses but no DC = Output amp) 4. Take a look at Section 6.5 OF the service manual and table 6.1 which gives suggestions in each of the circumstances I described. In terms of separate sections you can eliminate them by finding the components at the boundaries between sections and either checking there is a signal present or disconnecting and injecting a signal into the following stage. I can't give every possibility but here are a few examples: With it set to a few hundred KHz constant wavefrm 1. Remove Q44 (trigger output that stops the current switch), if it starts working problem is trigger/logic. 2. R122 is input to the preamp, so pull out one leg and inject a small signal 3. R137 is at the input to the output amp, so again pull up one leg and inject a small AC signal 4. Check GCV output and vary frequency knob, that will verify if the Current source (and perhaps AGC) are working 5 Check signal on R72, this is the Hysteresis/current switch interface The main board schematic is divided into the various sections - so try and narrow the sections then section 6 does give a few handy component level testing tips, Hope that helps
@@MakeOrRepairThank you for all that information. You've given me a lot to go on so I'll get it back on the bench and have another look. I do have the service manual which looks just like the one you used in this video so I'll be able to follow along with what you described.
@@MakeOrRepair Ok, lots of info for an update. Years ago when the 184 quit (working fine one day, dead the next day) I was aware of certain settings like the manual trigger keeping the output flatlined so I tried every setting with no changes. And I have been through section 6.5 troubleshooting numerous times. My test settings today are the x1Khz range (so 1-5Khz on the dial), freq. vernier all the way up, normal symmetry, generator mode continuous, triangle waveform without offset, amplitude at max with -0dB. In my past notes from previous times I've worked on it I noted the opamps at IC11 & 12 being bad which sent the ±15V lines to about +22V/-4V. Replacing both opamps brought back the ±15V lines. I noted I had a sweep output but no waveform output (I didn't test the sweep today). Previously the GCV out showed nothing so I replaced the LM301 opamp at IC2 and that returned the variable GCV output. At the time I tested the CA3096 transistor array at IC3 and noted it tested all good (I probably used my component tester to check one transistor at a time). All transistors tested good by pulling from the board and checking separately. The SN7400 NAND gate at IC7 was swapped with a new one with no change so I concluded it was good. Apparently I didn't check all the ICs at the time, either because I couldn't or had no replacements to swap. I thought I'd checked all the ICs back then but it seems there are several I don't have replacements for. With today's testing the GCV output does range from 0-5.2VDC, variable with the frequency dial. The TTL pulse output is stuck at 4.05VDC, and the 50Ω output is stuck at -11.8VDC (it does drop accordingly with higher attenuation settings on the amplitude knob). Regarding the checks you listed: 1. The ±15V & +5V lines test good; the +10V line shows +9.0V and the -5V line shows -5.2V; I don't know whether those are far enough out of spec to kill the thing. 2. As listed above, all the settings are correct and should be producing an output. All the knobs are tight (sweep board knobs are removed from when I pulled it; the board is back in place for testing). 3. There is NO pulse at the pulse output. 4. As I described above, I have been through section 6.5 many times so that's definitely been checked. Regarding the tests you suggested: 1. Removing Q44 makes no change. 2. Injecting a signal on R122 does show up at the 50Ω output, so it seems the preamp circuit is working. 3. Injecting a signal on R137 does show up on the 50Ω output, so it seems the output amp is working too. 4. Yes, the GCV output does vary, 0-5.2VDC. 5. The only signal on R72 is -12.47VDC. I can't find any oscillation whatsoever in the current switch circuit, hysteresis circuit, or the TTL pulse circuit (probably because it's getting nothing from hysteresis). I checked every leg of the CA3054 differential amplifier at IC4 in the hysteresis circuit; anything coming from +15 shows around +15 and anything coming from -15 shows around -15 (things vary based on resistors in the path). Where I would expect there to be a waveform coming from pin 4/Q20 collector (square wave 1 on the schematic) and pin 12/Q21 collector (square wave 2 on the schematic) there was either nothing or a flat DC voltage (I forgot to write that down). I'm starting to suspect IC4 may be bad but I have no way to test it, so I might just have to order a couple replacements to swap out and see. If the +15 line going up to 22V potentially fried IC2 then it may have fried IC4 as well. I can't find a good datasheet for it to know what the absolute maximum ratings for it are. And I just noticed another LM301 in the cap multiplier at IC5; I'll swap that out the next time I get back to the bench just in case that's shot too. It might be possible that the overvoltage going through the +10/-5V regulator resulted in a downstream overvolt that took out CA3130 at IC6 in the cap multiplier, but I don't have any spares of those to check either. On a side question, in the current switch, transistors Q9, 10, 11 and 12 all have absolutely nothing connected to the emitters. They're all soldered to a pad with no traces coming off it. Why is that and how can they do anything without that connection? Thanks again for all your help.
Before I start, I don't have a unit anymore so I can only look at the schematic and give advice how I would go about it... First, The transistors are being used a low current switches, the Signal on q13 that comes from the hysteresis switch is forcing opposite pairs into conduction or otherwise and causing the capacitor that is connected between q9 and q11 to alternatively charge and discharge via the two current sources giving the square wave that is dependent on the + and - current source. Not a method used that regularly although I think both Agilent and Tektronix have used in their high end multimeters. Your test so far are good news and bad since you have eliminated all the analogue amps and removed the trigger and know the VGC does something that is not to say there is no fault at all in those sections, just they are not preventing the oscillation. Now since there is no output pulse and you are missing pulses everywhere, the problem should lie somewhere in the control loop so current source, current switch, buffer amp or hysteresis switch. (note there may be a problem in the sync pulse generator stopping output pulse, but you can see several other places that pulses should occur!. It can always be difficult to troubleshoot circuits like this where there is a cyclic dependency, often the trick is to break the cycle. I would be tempted to start at the current sources and the current switch and break the cycle by unsoldering one end of R50 (the feedback from hysteresis to current switch) so it stays connected to the transistor Q13 in the current switch. You can now inject a switching squarewave signal into to dangling end of the resistor (not sure of amplitude but should go positive and negative (you can use a manual wire with say a 1k resistor and connect to +15 then -15 - switch the range switch to 1kHz (1kHz does not use the cap multiplier!) and monitor the voltage the junction of CR6 and CR7 (or the pole of the switch or anywhere convenient) fingers crossed you will see the capacitor charge and discharge as your input signal changes. If that doesn't work then either the switch is bad or the current sources. If it does seem to chage and discharge then progress forward and test the output of the N amp at the juntion of R55 and Q19 (marked N1 on the diagram. You should also be able to see some of the signal on pin2 of IC4 and then on the bases of q21 and q20. To test the cap multiplier switch to the 10Hz range and repeat! After that you will hopefully have finally narrowed the failure to oscilate down quite a lot and it is then a case of checking the components in the dicey section. I would say that in general a +22-5V on the opamps etc is unlikely to cause a problem, I suspect somehting far more nasty took place putting much higher voltages across the circuit and the opamps in the power supply were just one casualty. The fiirst thing however is to get that oscillation working then you can double-check the sine is nice and the logic sections like square way sharing and sync. Nothing much else is coming to mind, I can't really advise much on checking within the individual sections because I and just looking at the schamatic but hopefully I have helped you narrow it down.
@@MakeOrRepair Once again, thank you for all the information. I'll get back to it next weekend and see what I'm able to figure out with it. Hopefully I'll be able to avoid the brute force method of starting to just replace everything in those sections.
Sorry about sound, Using power director to remove some background noise and it was fine in preview, not so much after publishing - Soooory
Great video, very clear and understandable presentation. Bought a a Wavetek 184 today!
Thanks
Nice repair.
Thanks
I wish I could send my 184 to you for a fix. I've been working on it off and on for 4 years and can't figure out what's wrong. There's no signal output at the 50Ω out and tracing backwards on the board or the schematic I can't even figure out where it's supposed to originate. The only thing wrong I've ever found was the opamps in the power supply circuit were bad, so replacing them brought back the ±15V. Every transistor tests good, every cap tests good, there are no shorted resistors (the 3.9MΩ R14 in the VCG amplifier tests at 4.4MΩ but I don't think that would cause the problem), and every IC I've been able to test by swapping with a new one seems good. Any ideas?
Take a look at service manual, there is a block diageram before section 4. and the circuit diagrams in section 7 and board layouts too) are broken down into the mmatching logical blocks which helps isolate various parts for testing.
Everything starts with positive and negative current sources (which are set by the VCG amp) and a current switch alternates between them based on timing and the hysteresis circuit, now constant current into a timing capacitor gives a linear slope and since it alternates +/- it gives a triangle, the speed of the rise/fall depending on the amount of current. This triangle goes to a buffer amplifier (from which we get a triangle) and this then feeds back to the hysteresis switch, which in turn triggers the swapping of the currents again so the slope goes the other way.
*The TTL pulse output is taken from the Hysteresis section of the circuit so misses out the pre-amp and the output amplifier - so this is a handy place to check if you have a signal that is not getting amplified, or no oscillation at all.
* The triangle from the buffer amp goes to the waveform switch (for triangle) and the sine converter witch sends a sine wave to the switch) The square wave and pulse are connected to the waveform switch come from a square wave shaper which is in turn from the hysteresis switch, but the switch has a DC position so that handily bypasses almost everything and lets you check the output amplifier in isolation!
In terms of further fault finding it is all about eliminating whole sections, but first make certain you have all 5 voltages present - you say you ahve +-15 but check this is getting to the op amps e.g. IC 11 and 12
1. As well as +-15V check the +5, +10 and -5V supplies. The 5V uses a 7805 Regulator and can be checked on the logic ICs. The +10V, -5V are generated using a divider network and appear on the emitters of Q22 and Q23 so there are just 3 resistors and 2 transistors if those voltages are not present. you should also sheck the voltages reach IC6 ins 7 and 4, it might be a round can, if the board is positioned so that the 15V power supply is top right then the two transistors and the IC6 are at the bottom right corner.
2. Make sure the switch tops are not rotated on sharft and choose a constant signal that does not require any external trigger - easy fault to miss!
3. Check the pulse out - if it is present then the fault must be in the pre-amp or later. If it is present then try a DC waveform and vary the amplitude and see if that appears on output (no pulse = something before preamp, pulses and DC voltages = preamp, Pulses but no DC = Output amp)
4. Take a look at Section 6.5 OF the service manual and table 6.1 which gives suggestions in each of the circumstances I described.
In terms of separate sections you can eliminate them by finding the components at the boundaries between sections and either checking there is a signal present or disconnecting and injecting a signal into the following stage. I can't give every possibility but here are a few examples:
With it set to a few hundred KHz constant wavefrm
1. Remove Q44 (trigger output that stops the current switch), if it starts working problem is trigger/logic.
2. R122 is input to the preamp, so pull out one leg and inject a small signal
3. R137 is at the input to the output amp, so again pull up one leg and inject a small AC signal
4. Check GCV output and vary frequency knob, that will verify if the Current source (and perhaps AGC) are working
5 Check signal on R72, this is the Hysteresis/current switch interface
The main board schematic is divided into the various sections - so try and narrow the sections then section 6 does give a few handy component level testing tips,
Hope that helps
@@MakeOrRepairThank you for all that information. You've given me a lot to go on so I'll get it back on the bench and have another look. I do have the service manual which looks just like the one you used in this video so I'll be able to follow along with what you described.
@@MakeOrRepair Ok, lots of info for an update. Years ago when the 184 quit (working fine one day, dead the next day) I was aware of certain settings like the manual trigger keeping the output flatlined so I tried every setting with no changes. And I have been through section 6.5 troubleshooting numerous times. My test settings today are the x1Khz range (so 1-5Khz on the dial), freq. vernier all the way up, normal symmetry, generator mode continuous, triangle waveform without offset, amplitude at max with -0dB.
In my past notes from previous times I've worked on it I noted the opamps at IC11 & 12 being bad which sent the ±15V lines to about +22V/-4V. Replacing both opamps brought back the ±15V lines. I noted I had a sweep output but no waveform output (I didn't test the sweep today). Previously the GCV out showed nothing so I replaced the LM301 opamp at IC2 and that returned the variable GCV output. At the time I tested the CA3096 transistor array at IC3 and noted it tested all good (I probably used my component tester to check one transistor at a time). All transistors tested good by pulling from the board and checking separately. The SN7400 NAND gate at IC7 was swapped with a new one with no change so I concluded it was good. Apparently I didn't check all the ICs at the time, either because I couldn't or had no replacements to swap. I thought I'd checked all the ICs back then but it seems there are several I don't have replacements for.
With today's testing the GCV output does range from 0-5.2VDC, variable with the frequency dial. The TTL pulse output is stuck at 4.05VDC, and the 50Ω output is stuck at -11.8VDC (it does drop accordingly with higher attenuation settings on the amplitude knob). Regarding the checks you listed:
1. The ±15V & +5V lines test good; the +10V line shows +9.0V and the -5V line shows -5.2V; I don't know whether those are far enough out of spec to kill the thing.
2. As listed above, all the settings are correct and should be producing an output. All the knobs are tight (sweep board knobs are removed from when I pulled it; the board is back in place for testing).
3. There is NO pulse at the pulse output.
4. As I described above, I have been through section 6.5 many times so that's definitely been checked.
Regarding the tests you suggested:
1. Removing Q44 makes no change.
2. Injecting a signal on R122 does show up at the 50Ω output, so it seems the preamp circuit is working.
3. Injecting a signal on R137 does show up on the 50Ω output, so it seems the output amp is working too.
4. Yes, the GCV output does vary, 0-5.2VDC.
5. The only signal on R72 is -12.47VDC.
I can't find any oscillation whatsoever in the current switch circuit, hysteresis circuit, or the TTL pulse circuit (probably because it's getting nothing from hysteresis). I checked every leg of the CA3054 differential amplifier at IC4 in the hysteresis circuit; anything coming from +15 shows around +15 and anything coming from -15 shows around -15 (things vary based on resistors in the path). Where I would expect there to be a waveform coming from pin 4/Q20 collector (square wave 1 on the schematic) and pin 12/Q21 collector (square wave 2 on the schematic) there was either nothing or a flat DC voltage (I forgot to write that down). I'm starting to suspect IC4 may be bad but I have no way to test it, so I might just have to order a couple replacements to swap out and see. If the +15 line going up to 22V potentially fried IC2 then it may have fried IC4 as well. I can't find a good datasheet for it to know what the absolute maximum ratings for it are. And I just noticed another LM301 in the cap multiplier at IC5; I'll swap that out the next time I get back to the bench just in case that's shot too. It might be possible that the overvoltage going through the +10/-5V regulator resulted in a downstream overvolt that took out CA3130 at IC6 in the cap multiplier, but I don't have any spares of those to check either.
On a side question, in the current switch, transistors Q9, 10, 11 and 12 all have absolutely nothing connected to the emitters. They're all soldered to a pad with no traces coming off it. Why is that and how can they do anything without that connection?
Thanks again for all your help.
Before I start, I don't have a unit anymore so I can only look at the schematic and give advice how I would go about it...
First, The transistors are being used a low current switches, the Signal on q13 that comes from the hysteresis switch is forcing opposite pairs into conduction or otherwise and causing the capacitor that is connected between q9 and q11 to alternatively charge and discharge via the two current sources giving the square wave that is dependent on the + and - current source. Not a method used that regularly although I think both Agilent and Tektronix have used in their high end multimeters.
Your test so far are good news and bad since you have eliminated all the analogue amps and removed the trigger and know the VGC does something that is not to say there is no fault at all in those sections, just they are not preventing the oscillation. Now since there is no output pulse and you are missing pulses everywhere, the problem should lie somewhere in the control loop so current source, current switch, buffer amp or hysteresis switch. (note there may be a problem in the sync pulse generator stopping output pulse, but you can see several other places that pulses should occur!.
It can always be difficult to troubleshoot circuits like this where there is a cyclic dependency, often the trick is to break the cycle. I would be tempted to start at the current sources and the current switch and break the cycle by unsoldering one end of R50 (the feedback from hysteresis to current switch) so it stays connected to the transistor Q13 in the current switch. You can now inject a switching squarewave signal into to dangling end of the resistor (not sure of amplitude but should go positive and negative (you can use a manual wire with say a 1k resistor and connect to +15 then -15 - switch the range switch to 1kHz (1kHz does not use the cap multiplier!) and monitor the voltage the junction of CR6 and CR7 (or the pole of the switch or anywhere convenient) fingers crossed you will see the capacitor charge and discharge as your input signal changes. If that doesn't work then either the switch is bad or the current sources. If it does seem to chage and discharge then progress forward and test the output of the N amp at the juntion of R55 and Q19 (marked N1 on the diagram. You should also be able to see some of the signal on pin2 of IC4 and then on the bases of q21 and q20. To test the cap multiplier switch to the 10Hz range and repeat!
After that you will hopefully have finally narrowed the failure to oscilate down quite a lot and it is then a case of checking the components in the dicey section. I would say that in general a +22-5V on the opamps etc is unlikely to cause a problem, I suspect somehting far more nasty took place putting much higher voltages across the circuit and the opamps in the power supply were just one casualty.
The fiirst thing however is to get that oscillation working then you can double-check the sine is nice and the logic sections like square way sharing and sync.
Nothing much else is coming to mind, I can't really advise much on checking within the individual sections because I and just looking at the schamatic but hopefully I have helped you narrow it down.
@@MakeOrRepair Once again, thank you for all the information. I'll get back to it next weekend and see what I'm able to figure out with it. Hopefully I'll be able to avoid the brute force method of starting to just replace everything in those sections.