Way back in the 80's I built a power amp, I think it is an ETI5000. It has worked like a dream all this time except for (rare) blown fuses when I played the Telarc recording of Tchaikovski's 1812 at full blast (the cannon shots did it).
Recently the amp started playing-up (after 20 yrs of good service). One channel drops out after about 5 min of music. I need to troubleshoot but I can't find the original ETI articles with the circuit diagram and circuit description (I think there were two articles).
Can anyone help?
Recently the amp started playing-up (after 20 yrs of good service). One channel drops out after about 5 min of music. I need to troubleshoot but I can't find the original ETI articles with the circuit diagram and circuit description (I think there were two articles).
Can anyone help?
i think you should email silicon chip and see if they can get a hold of the article, i think they have archives of all the old EA and ETI stuff. at the most i think they chare 8.80 for a back issue article.
however a google image search for eti 5000 found this image
looks like a schematic for the amplifier in question....
however a google image search for eti 5000 found this image
looks like a schematic for the amplifier in question....
An externally hosted image should be here but it was not working when we last tested it.
ETI 5000 amp problem
Thanks heaps for the advice and the schematic. I should be able to fix it now, especially since I have one channel working so I can make comparisons.
My (not very well educated) guess is that one of the transistors driving the FETs has an age-related thermal runaway problem. Thank goodness my age related problems do not include that!
If anyone out there can offer advice on what to look for when troubleshooting, it would be much appreciated.
Alex
Thanks heaps for the advice and the schematic. I should be able to fix it now, especially since I have one channel working so I can make comparisons.
My (not very well educated) guess is that one of the transistors driving the FETs has an age-related thermal runaway problem. Thank goodness my age related problems do not include that!
If anyone out there can offer advice on what to look for when troubleshooting, it would be much appreciated.
Alex
It's probably worth changing all electrolytic capacitors as well, there could be a problem with the input coupling cap (C1) causing this. Go for a polyester unit if you can. I would guess the power supply capacitors could also do with replacing.
I would have thought getting BF469/470 would be a problem but no - Jaycar has them!
I would have thought getting BF469/470 would be a problem but no - Jaycar has them!
I have repaired mine a few times but it has not been used for some time now. It is often easier to replace all the low level and driver transistors, saves trying to find the exact transistor that is causing the problem.
Also the heat sink for the bf469 bf470 transistors is very small in the original design. I have used much larger aluminium angles and connected that to a horizontal piece of aluminium that spans both boards with the cross piece of aluminium extending beyond the driver transistors.
This amp can also oscillate due to the mosfet 5w resistors being slightly inductive. One solution was to connect a capacitor across to join the 2 mosfets together (220nf from memory) or replce with parallel 1w carbon films to make them up to 5w or 6w power dissipation.
Regards,
Wayne
Also the heat sink for the bf469 bf470 transistors is very small in the original design. I have used much larger aluminium angles and connected that to a horizontal piece of aluminium that spans both boards with the cross piece of aluminium extending beyond the driver transistors.
This amp can also oscillate due to the mosfet 5w resistors being slightly inductive. One solution was to connect a capacitor across to join the 2 mosfets together (220nf from memory) or replce with parallel 1w carbon films to make them up to 5w or 6w power dissipation.
Regards,
Wayne
ETI 5000 Problem
Alex
Replace C5,C12 and C13 in both channels. They dry out after a while and may cause instability problems and increased distortion.
Wayne is correct about the need for extra heatsinking.
Also, check R29 for discolouration due to overheating caused by instability. The thread at the link below may be of interest.
SandyK (another Alex !)
http://www.diyaudio.com/forums/showthread/t-103649.html
Alex
Replace C5,C12 and C13 in both channels. They dry out after a while and may cause instability problems and increased distortion.
Wayne is correct about the need for extra heatsinking.
Also, check R29 for discolouration due to overheating caused by instability. The thread at the link below may be of interest.
SandyK (another Alex !)
http://www.diyaudio.com/forums/showthread/t-103649.html
I repaired my dad's one a couple of months ago.
The symptom was left channel very small output, only 1/2-wave driven, very quiet and distorted. on turn-on, the output swung momentarily towards -ve rail, then stabilised at zero. Right channel was fine.
one of the driver transistors had died (Q6 IIRC) causing the driver stage to go fully -ve, and the dc offset corrected it back to zero, leaving only a little bit of rail swing for the output stage.
anyways, I replaced all the transistors in the driver stage (Q3 to Q8) cost me about $20, and replaced all the insulating washers and thermal paste at the same time. I thought I felt a burr on the heatsink for the 6 transistors, so I made the sink all flat too - just to be sure.
reassembled and test - very nice sounding amp.
The symptom was left channel very small output, only 1/2-wave driven, very quiet and distorted. on turn-on, the output swung momentarily towards -ve rail, then stabilised at zero. Right channel was fine.
one of the driver transistors had died (Q6 IIRC) causing the driver stage to go fully -ve, and the dc offset corrected it back to zero, leaving only a little bit of rail swing for the output stage.
anyways, I replaced all the transistors in the driver stage (Q3 to Q8) cost me about $20, and replaced all the insulating washers and thermal paste at the same time. I thought I felt a burr on the heatsink for the 6 transistors, so I made the sink all flat too - just to be sure.
reassembled and test - very nice sounding amp.
ETI 5000 amp problem
Wow, thanks you guys (sorry, people). Good to know the transistors can still be sourced (thanks, Jaycee. Pulse-R). I will replace the low power transistors and if that doesn't work, turn to the capacitors (thanks WayneK, SandyK and Pulse-R).
I really do appreciate your help. It's great to know that there are people out there willing to help! Kinda restores my faith in humanity.
The ETI 5000 is a great amplifier. Mine is connected to four speaker boxes (two channels, each two boxes). Boxes are 8 Ohm so the load is 4 Ohms per channel for 100W per channel output from the ETI 5000.
Speakers are Philips 12 inch woofers (20 Hz cone resonance) and 5 inch midrange in each box with four Audax tweeters (series/parallel to get 8 Ohm from 4x8 Ohm). The boxes are big (each 4x2x1 foot approx) and totally sealed so the low freq response is superb! Clean but able to make the whole (weatherboard) house shake.
Wow, thanks you guys (sorry, people). Good to know the transistors can still be sourced (thanks, Jaycee. Pulse-R). I will replace the low power transistors and if that doesn't work, turn to the capacitors (thanks WayneK, SandyK and Pulse-R).
I really do appreciate your help. It's great to know that there are people out there willing to help! Kinda restores my faith in humanity.
The ETI 5000 is a great amplifier. Mine is connected to four speaker boxes (two channels, each two boxes). Boxes are 8 Ohm so the load is 4 Ohms per channel for 100W per channel output from the ETI 5000.
Speakers are Philips 12 inch woofers (20 Hz cone resonance) and 5 inch midrange in each box with four Audax tweeters (series/parallel to get 8 Ohm from 4x8 Ohm). The boxes are big (each 4x2x1 foot approx) and totally sealed so the low freq response is superb! Clean but able to make the whole (weatherboard) house shake.
sandyK
my understanding of the 5000's driver stage, is that the DC feedback will serve to correct the output's DC offset by biasing the transistors on/off more or less to maintain zero output. It seemed that way when I was testing, IIRC it was Q4 stuck 'on' causing the problem - on power-up, the output would swing down near V- and then, over about 0.5 seconds, the output would return to zero. with the driver stages stuck hard, there was only very small amount of 'swing' left, so the signal was easily distorted. Maybe I just imagined it that way, but my theory turned out well in practice.
my understanding of the 5000's driver stage, is that the DC feedback will serve to correct the output's DC offset by biasing the transistors on/off more or less to maintain zero output. It seemed that way when I was testing, IIRC it was Q4 stuck 'on' causing the problem - on power-up, the output would swing down near V- and then, over about 0.5 seconds, the output would return to zero. with the driver stages stuck hard, there was only very small amount of 'swing' left, so the signal was easily distorted. Maybe I just imagined it that way, but my theory turned out well in practice.
ETI 5000 amp problem
Hi SandyK,
Yes, my recollection is that negative feedback is used to set the overall DC gain to 1. So it would make sense that if major DC drift occurs the feedback might reduce the available swing to near zero. Good heavens (I was going to say 'bugger' but that is not polite) it has been 20 years since I built the thing - no wonder my recollections are bad. Anyway, I have faith (?) that replacing all the low level transistors should fix it.
Thanks,
Alex
Hi SandyK,
Yes, my recollection is that negative feedback is used to set the overall DC gain to 1. So it would make sense that if major DC drift occurs the feedback might reduce the available swing to near zero. Good heavens (I was going to say 'bugger' but that is not polite) it has been 20 years since I built the thing - no wonder my recollections are bad. Anyway, I have faith (?) that replacing all the low level transistors should fix it.
Thanks,
Alex
ETI 5000 Problem
Pulse-R and AlexRWard
Hey guys, my recollection is hazy too ! Mine was built in 1981. (?)
I only asked the question because of the wording "and the dc offset corrected it back to zero," I wrongly jumped to the conclusion that yoy may have been talking about the 6000, which had a small DC offset corrector PCB fitted.
If you are interested in the small modification mentioned in that other thread, and discussed by Suzy, send me an email and I will reply with an underneath photo attached. It only involves cutting a single track and connecting 2 series connected 9V zeners across it. A 18V 1W zener would also be suitable.
SandyK
Pulse-R and AlexRWard
Hey guys, my recollection is hazy too ! Mine was built in 1981. (?)
I only asked the question because of the wording "and the dc offset corrected it back to zero," I wrongly jumped to the conclusion that yoy may have been talking about the 6000, which had a small DC offset corrector PCB fitted.
If you are interested in the small modification mentioned in that other thread, and discussed by Suzy, send me an email and I will reply with an underneath photo attached. It only involves cutting a single track and connecting 2 series connected 9V zeners across it. A 18V 1W zener would also be suitable.
SandyK
interesting you mention the mods, and I had no idea at the time -
this one had the zener diodes, and extra capacitors on the output transistors... hmmm, I guess the old man wasn't too slow in the olden days
Might have to look into the SuzyJ amp there, might be another 'must-do' to add to the list.
this one had the zener diodes, and extra capacitors on the output transistors... hmmm, I guess the old man wasn't too slow in the olden days
Might have to look into the SuzyJ amp there, might be another 'must-do' to add to the list.
ETI 5000 Problem
Pulse-R
In my case, the mod improved sound quality as it reduced the uneven loads on both halves of the previous stage. The mod stopped one half of the differential pair from getting markedly hotter than the other side . This meant that their VBEs were now similar, and less unbalance of the previous stage. The side where the zener was inserted, previously had more voltage dissipated in it's transistor, and thus due to being hotter, a lower VBE than the other side.
The mod was discussed with David Tilbrook at Jaycar's Gore Hill store and approved by him.
SandyK
Pulse-R
In my case, the mod improved sound quality as it reduced the uneven loads on both halves of the previous stage. The mod stopped one half of the differential pair from getting markedly hotter than the other side . This meant that their VBEs were now similar, and less unbalance of the previous stage. The side where the zener was inserted, previously had more voltage dissipated in it's transistor, and thus due to being hotter, a lower VBE than the other side.
The mod was discussed with David Tilbrook at Jaycar's Gore Hill store and approved by him.
SandyK
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.