your version is much better I guess. What transistor are they? 2n5551 is not to-220, it doesn't have a heat sink.
I might upgrade it to to-220, but what model?
Thanks
Unfortunately, my amp didn't sound particularly good. Perhaps this was because the power transistors fitted were 2SC5200/A1943 - probably the only replacements that were available in volume here in Oz after failure. That could mean many things but whatever repairs were done, the results weren't acceptable, IMHO.
Really, any NPN transistor will work as a bias/Vbe multiplier transistor but the pinout configuration should suit the PCB, which likely was designed for EBC pinout if specified 2N5551 and reading normally from the label side. To substitute, you'd need something with the same pinout but that isn't easy with common TO220 or TO126 transistor types that are all ECB or BCE respectively.
So, you still require US style TO92 transistors (again, assuming they are original types and pinout). I've noticed several Chinese 2N marked parts with Japanese pinouts in recent years - an unregulated, undocumented situation in the Asian semiconductor market that causes mayhem. Having a significantly different bias current in either channel might indicate something odd there.
Presumably, your 2N5551 transistors are clamped with the flat side to the heatsink and that's fine - they don't require much thermal coupling in their temperature sensing mode. There's nothing more you need do there if they are functional and bias is acceptably close to spec.
Really, any NPN transistor will work as a bias/Vbe multiplier transistor but the pinout configuration should suit the PCB, which likely was designed for EBC pinout if specified 2N5551 and reading normally from the label side. To substitute, you'd need something with the same pinout but that isn't easy with common TO220 or TO126 transistor types that are all ECB or BCE respectively.
So, you still require US style TO92 transistors (again, assuming they are original types and pinout). I've noticed several Chinese 2N marked parts with Japanese pinouts in recent years - an unregulated, undocumented situation in the Asian semiconductor market that causes mayhem. Having a significantly different bias current in either channel might indicate something odd there.
Presumably, your 2N5551 transistors are clamped with the flat side to the heatsink and that's fine - they don't require much thermal coupling in their temperature sensing mode. There's nothing more you need do there if they are functional and bias is acceptably close to spec.
Thanks Ian,
I always trust RS component with parts, but they sell these transistors in buld of 100.
So, I tried somewhere else, with hFE test on my multimeter, hopefully I can get it done.
Will replace both in case they are on different specs.
Yes, RS is a good local supplier but I was surprised to find US based suppliers can freight complete orders from Asia to my door in as little as 4 working days, freight-free above a $50-60 threshold! UK based suppliers RS and Element 14 tend to footle about with split deliveries coming from stock that is actually held in multiple locations around the globe. My last order of 6 items took a month to finalise.
As mentioned, you don't specifically need 2N5551 here and you can buy 1-off quantities of MPSA42, MPSA06 for example, fairly cheap. (Jaycar) Any US type NPN of Vceo 80V and above will probably do. You don't even need to check the Hfe as that's not an issue but having the right pinout is essential. One of those cheap Ebay component testers comes into its own with identifying the leads but you can still do that with a multimeter, following a transistor testing guide on just about any Googled electronics teaching website.
As mentioned, you don't specifically need 2N5551 here and you can buy 1-off quantities of MPSA42, MPSA06 for example, fairly cheap. (Jaycar) Any US type NPN of Vceo 80V and above will probably do. You don't even need to check the Hfe as that's not an issue but having the right pinout is essential. One of those cheap Ebay component testers comes into its own with identifying the leads but you can still do that with a multimeter, following a transistor testing guide on just about any Googled electronics teaching website.
thanks Ian.
Actually, I think my amp sounds nice after the mod.
The problem is only with measuring. The question is, should I worry about the unstable DC output on the right channel(supposedly faulty)?
If you mean the +/- 40mV offset, it wouldn't be a problem if it was a gradual drift and thus traceable. Being only in one channel and cycling slowly between narrow limits says there is a fault with a time constant in the DC feedback to the input stage and this is the worrying part. It could be thermal or capacitive in origin. I would have suspected any electrolytic that is close to a heat source but I'm not sure that's the problem here. It could also be associated with the input stage.
Before going further though, ground the input to the power amplifier stage at the wiper of RP101 to ensure there are fewer external influences. I would next look for any coincidental voltage fluctuation, starting with say, across C228, 233, 234, 240. Resolder these and other connections in the area and retest, attaching the meter leads firmly with clips where you can or using sharp point probes with enough pressure to ensure the measurements at least, are stable.
The feedback that controls DC is summed from both output transistor bases so if there is instability between the bases there, it's possible there would be an incremental +/- change in offset, so look carefully for a source of that switching polarity when there is no signal or power supply fluctuation to cause other disturbances. 40mV or possibly a lot less, is not much to look for.
Before going further though, ground the input to the power amplifier stage at the wiper of RP101 to ensure there are fewer external influences. I would next look for any coincidental voltage fluctuation, starting with say, across C228, 233, 234, 240. Resolder these and other connections in the area and retest, attaching the meter leads firmly with clips where you can or using sharp point probes with enough pressure to ensure the measurements at least, are stable.
The feedback that controls DC is summed from both output transistor bases so if there is instability between the bases there, it's possible there would be an incremental +/- change in offset, so look carefully for a source of that switching polarity when there is no signal or power supply fluctuation to cause other disturbances. 40mV or possibly a lot less, is not much to look for.
Thanks.
It was changing rapidly between 40mv and -40mv, without any circling. Right speaker had noise.
Interestingly, after I desoldered those capacitor to check and found no issues, then re-solder them on, now the DC output has dramatically dropped to +-2mv. And there noise from the right speaker is almost not audible even if I stay very close, say 5cm away from them, along with slight humming from both.
Thank you very much. This is pretty good result, I reckon, for such a cheap amp.
Last edited:
unfortunately, after playing for a day without problem, after above mentioned check done,
now the bias has returned to zero on right channel. 🙁
now the bias has returned to zero on right channel. 🙁
I think that's an intermittent fault such as bad soldering, faulty component etc, if you can make bias current disappear and reappear by pulling and refitting unrelated components. How are you measuring this bias? It should be measured as the voltage in mV across the emitter resistor of either output transistor. By Ohm's law, you then calculate current if that is the specified format.
How would you describe the noise in #27 - Hiss? High pitched? Low? random crackling?
How would you describe the noise in #27 - Hiss? High pitched? Low? random crackling?
The bias is measured across two emitter resistors. About 34mv in ideal situation, which makes Iq=121mA. (4*0.39R and 1*1R.)
It was random crackling, very low though.
After one night, the bias is back. 🙁
Hi,
Sometime when I am troubleshoot an amplifier with an intermittent problem just take a blower and try to heat by small area and if you have an intermittent component it will fail. The heat will make the components failed if it is an intermittent component. Just a suggestion.
Sometime when I am troubleshoot an amplifier with an intermittent problem just take a blower and try to heat by small area and if you have an intermittent component it will fail. The heat will make the components failed if it is an intermittent component. Just a suggestion.
Hmm.. Random crackling is often a sign of instability - oscillation. It may not be at a high level but it's most commonly at RF frequencies below 500 kHz, so you can't hear it directly but some low level noise like sharpish hiss and crackling is evidence of it.
It really needs a look-see with a decent oscilloscope. Just a comparison of the outputs from each channel would identify that sort of problem. It's not clear by other methods but before I could afford a 'scope that didn't need a crane to lift it, I usually could pick up the noise on a pocket AM radio, tuned about at the bottom of the dial and waved about over the output stage. Harmonics of the oscillation should be evident at higher frequencies too, if you can't tune low enough. I dare say a long wire coiled up and used as the AM aerial lead on a AM/FM tuner would also work but they're usually more immune to out-of-band noise.
I recall that by removing power alternately from each channel, the sound was like several rough tones that shifted about as I tried to home in on it but that's not really necessary - just that an unusual noise comes and goes with power to the amplifier - or not. There are other simple indicator gizmos using leds that can be used by connecting to the OPS. For example, this one by past member Bonsai: http://hifisonix.com/wordpress/wp-content/uploads/2017/10/Oscillation-Sniffer-by-Bonsai.pdf
Instability could be due to many causes and as I mentioned , these amps were problematic and many were returned with replaced output semis and other mods around the OPS. They have no speaker protection relay and use a current limiting circuit to reduce excessive drive to the output stage (V226,227,228,230). This type of limiter has been around as long as solid state but was outmoded by the 1980s. It's very fast, silent, effective and cheap but it doesn't save the speakers in a catastrophic meltdown and can bring other problems if not done right.
It really needs a look-see with a decent oscilloscope. Just a comparison of the outputs from each channel would identify that sort of problem. It's not clear by other methods but before I could afford a 'scope that didn't need a crane to lift it, I usually could pick up the noise on a pocket AM radio, tuned about at the bottom of the dial and waved about over the output stage. Harmonics of the oscillation should be evident at higher frequencies too, if you can't tune low enough. I dare say a long wire coiled up and used as the AM aerial lead on a AM/FM tuner would also work but they're usually more immune to out-of-band noise.
I recall that by removing power alternately from each channel, the sound was like several rough tones that shifted about as I tried to home in on it but that's not really necessary - just that an unusual noise comes and goes with power to the amplifier - or not. There are other simple indicator gizmos using leds that can be used by connecting to the OPS. For example, this one by past member Bonsai: http://hifisonix.com/wordpress/wp-content/uploads/2017/10/Oscillation-Sniffer-by-Bonsai.pdf
Instability could be due to many causes and as I mentioned , these amps were problematic and many were returned with replaced output semis and other mods around the OPS. They have no speaker protection relay and use a current limiting circuit to reduce excessive drive to the output stage (V226,227,228,230). This type of limiter has been around as long as solid state but was outmoded by the 1980s. It's very fast, silent, effective and cheap but it doesn't save the speakers in a catastrophic meltdown and can bring other problems if not done right.
P.S.
You're right tauro, I should have thought of external heat as a way to check for an intermittent.
Re. using the amplifier under test: Don't try to take measurements while the speakers are connected. That will surely show unstable readings if there is noise at the output terminals of any kind - audible or not. If you have been, remove and recheck.
Bonsai's oscillation checker isn't the one I thought it was. I'll take another look around.....
This thread has some options that may even be within your junkbox parts range: Oscillation Sniffer
You're right tauro, I should have thought of external heat as a way to check for an intermittent.
Re. using the amplifier under test: Don't try to take measurements while the speakers are connected. That will surely show unstable readings if there is noise at the output terminals of any kind - audible or not. If you have been, remove and recheck.
Bonsai's oscillation checker isn't the one I thought it was. I'll take another look around.....
This thread has some options that may even be within your junkbox parts range: Oscillation Sniffer
Last edited:
Thanks for the elaborated explanation. Much appreciated.
Although it does sounds good on my testing epos shelf speakers already, I am hesitating to connect my Yamaha ns1000m speakers to it.
There is a speaker protection relay though, when the first 2sc5200 burned, the epos speaker survived.
My original idea is to use its power and chassis to build a CG, lm3886 specifically, as my previous one sounded very nice.
Maybe I should scrap it and build another CG?
Although it does sounds good on my testing epos shelf speakers already, I am hesitating to connect my Yamaha ns1000m speakers to it.
There is a speaker protection relay though, when the first 2sc5200 burned, the epos speaker survived.
My original idea is to use its power and chassis to build a CG, lm3886 specifically, as my previous one sounded very nice.
Maybe I should scrap it and build another CG?
'My mistake with the relay's DC protection - 'just couldn't see a sensing point but it's between the 10R Zobel resistor and cap. That doesn't seem wise. I did check out my amplifier after buying it and TBH, I don't recall all that much about it as it was quite some years ago. I didn't like it much and only took some interest when fellow Aussie, Greg Erskine, reworked the design and published details on his TPG /users website. I got the impression that it was a really worthwhile project. Mine worked OK but I still haven't touched it and it rests in my shed somewhere.
By way of personal experience, I've built many amplifiers based on magazine projects without any protection apart from rail fuses and even so, I've had the output transistors fail without taking the speakers with them. 'also worked on Japanese domestic amp. models that would fry anything connected to them and burn or weld their contacts together if the output stage went DC 😱
You probably used this site as a source for info. but you might look at the modifications too before deciding to scrap something that could turn out to be better than you may have expected. Come to think of it, you did mention "mod". Is your amp. now modified in some way? DSE A2760 Amplifier Modification - OCAU Wiki
By way of personal experience, I've built many amplifiers based on magazine projects without any protection apart from rail fuses and even so, I've had the output transistors fail without taking the speakers with them. 'also worked on Japanese domestic amp. models that would fry anything connected to them and burn or weld their contacts together if the output stage went DC 😱
You probably used this site as a source for info. but you might look at the modifications too before deciding to scrap something that could turn out to be better than you may have expected. Come to think of it, you did mention "mod". Is your amp. now modified in some way? DSE A2760 Amplifier Modification - OCAU Wiki
Yep. I saw this link but I followed diyaudio thread here somewhere to modify it as a dedicated power amp.
The mod in this link is somewhat partial, I reckon.
Maybe you are right, this amp appears to a inadequate design, and I dont like the quality of the parts and PCB.
Therefore, a CG is considered as cg has very good reviews.
- Status
- Not open for further replies.