I am waiting for a mouser order to arrive, should be here any day now. I most likely won't get much done before mid-december tho' as end of year school is both hectic and draining. I made up my mouser order with parts for a Mick Feuerbacher simple 3886 chipamp, but will have to get the Sansui right first. SWMBO is not impressed with multiple projects on the go.

We are what we pretend to be, so we must be careful about what we pretend to be.

Nov 1, 2020 14:55:01 GMT 12 papahemi said:

SWMBO is not impressed with multiple projects on the go.

You mean like this ?

darklantern.proboards.com/post/25391/thread

Just tell her multitasking.

It's our natural state, multiple states ...

Or desktops? Surely ...

It's the way(s) we get things done.

A rising tide floats all boats.

Cliche after cliche.

It could be a sign of insecurity to have to obey the other half all the time. I reckon have as many projects as you like!

AMR-iFi R&D

finally found the time - motivation to get back onto this today. I replaced all the electrolytic caps on poweramp board - it went quite smoothly. managed to reflow a few of the traces too:




all old caps seem quite sound bar one:



but it measures ok.

sorry bout the crap pics - i blame the crap photographer.

Next up the transistors and VRs
Then make a dim bulb tester before initial power up.
Hei ho

We are what we pretend to be, so we must be careful about what we pretend to be.

Success - what a great feeling when something goes right (almost). 

yesterday replaced all transistors on main board and 2x dodgy VRs, today i hooked up my dim bulb tester - no worries. Set current at 16ma - couldn't get much adjustment here - hence almost there.
Anyway plugged in some speakers and she went... so back into the main system and it does sound a lot better than the rotel pre and wee valve power amp - could be that there is a bit more grunt, but also more dynamic.
 Here she is back where she belongs:




Happy chappy - thanks to colinf for the pointers.

We are what we pretend to be, so we must be careful about what we pretend to be.

Hi Papahemi, excellent!
Which transistors did you end up using, and did you put them in homemade sockets?
The amp was most likely designed for class B, keep an eye (or hand) on the heatsink temperature to make sure all is well.

AMR-iFi R&D

i have used KSC1815 in place of 2SC281
ZTX453 instead of 2SC734
ZTX653 instead of CDC8002-1
ZTX753 instead of CDC9002-1

I did not end up using sockets - everything was able to fit in well if not a wee bit titaha. I did monitor heatsink temp for first hour or so, and upon checking now it does seem to be running quite hot - hand hot. 55.5c on my Blue Gizmo. Could this be associated with my inability to adjust current above 16ma?

We are what we pretend to be, so we must be careful about what we pretend to be.

That’s quite warm. The service manual states to remove the power rail fuse and measure the current, and set it to 22 to 26mA.

AMR-iFi R&D

Too warm, yesterday I opened it up again and tired to adjust bias current again with no joy. I did some reading on AK where a poster had similar issues and was advised to measure v across r839 and r840, adjusting to 13mv. This I did and is now runs cool. I don't really know what is going on here but afterwards I got more acceptable bias current readings. I think that the setting of vr901 (protection circuit) may be impacting here as I did move this in my initial search for problems. I cannot find a way to set this properly but I did return it to its more or less original position then was able to get the 13mv reading.
Still sounds good tho'

We are what we pretend to be, so we must be careful about what we pretend to be.

The protection circuit is quite nifty in its simplicity. SCR901 is an SCR that detects when there’s too much current being drawn from the speaker, triggers, and shorts out the driver transistor’s power supply in both channels.
VR901 adjusts the SCR’s trigger threshold. It triggers at about 0.6V on the gate (in theory anyway, the actual 2SF656 SCR might be slightly different. The datasheet says 0.8V max.). Current running through R839 (L) and R840 (R) is effectively in series with the speaker on negative half-cycles, and develops the voltage across it needed to trigger the SCR. With VR901 set to its most sensitive position the current needed through R839 or R840 is around 2.4A. With it set to mid position the current is 3.6A. With it set to short out the SCR gate the protection is switched off altogether, leaving just the rail fuses to protect the output stage.
Just briefly, here is how an SCR’s equivalent circuit looks:

AMR-iFi R&D

This is well over my head, but I take it that the setting on vr901 will have an impact, but how to know what the correct setting is? Meanwhile the Sansui continues to run cool, although I have now detected a low level buzz across both channels... not enough to detract from listening, but noticeable enough be tween tracks to have me worrying.

We are what we pretend to be, so we must be careful about what we pretend to be.

Ok, so you had difficulty setting the bias, there’s a low level buzz noticeable between tracks and adjusting VR901 affects the reading. I wonder if it’s oscillating. Do you have access to an oscilloscope? If so put it across the speaker terminals and see if all is quiet.

AMR-iFi R&D

no oscilloscope here. The buzz is very low and it runs cool. I have considered purchasing a cheap digital scope something like this:
www.trademe.co.nz/a/marketplace/business-farming-industry/industrial/electrical-equipment/gauges-meters/listing/2946071762?bof=NHrBwTU9
but I know little about how useful it would be.

We are what we pretend to be, so we must be careful about what we pretend to be.

That looks fine, great value for about $68! That’s possibly all you need for occasional use. Battery powered too, so it’s portable and less prone to mains-borne RF. I do all my low level measurements with a battery scope.
Don’t worry, I’ll teach you how to use it.

AMR-iFi R&D

But it only goes up to 200kHz!
Analog Bandwidth: 0-200KHz

A software implementation with added added display & etc. it seems.
There are software solutions for sound cards ... free, if no better bandwidth?, just FYI.

www.startpage.com/do/dsearch?query=free+software+oscilloscope

Old at least 10MHz+ scopes come up on TM from time to time.
I have a couple of old Eico 460 to restore, a working Solartron dual trace 10MHz, all valve, and in current use a SS Tektronics 2225 50MHz dual trace.
Still old, but enough at the moment.

See if you can find a local tech in the same city or area who has one, & can check for oscillation?
Simpler than buying such a cheap low value or performance scope, just to check that out, I'd think?

I'd offer here, but I'm in Auckland (and quite busy, almost overwhelmed at the moment!) and freight expensive.

Just thoughts in reaction. :-)

(I might be wrong, I might be right;-)

so what bandwidth would be needed, what frequency would the oscillation be? Maybe I will check with local tech if this turns out to be not suitable, but if it is I could have future use in the chipamp I am planning to build.

We are what we pretend to be, so we must be careful about what we pretend to be.

I've got a pretty good s/h 20MHz 'scope (Iwata) & it's good for up to 600V, but...

- For such SS work, you won't need high voltage,
- I've got 2 -3 'dead' scopes - ie old, complex equipment like this often becomes faulty either with heavy use or, from lack of use!
- These days, digital or even software-based scopes are so convenient & inexpensive, I'd consider one.

I'm no expert however

DarkLantern blog - darklanternforowen.wordpress.com

If you’re developing amplifiers, especially solid state, they can have a tendency to oscillate up to say 1Mhz, even more. The 200kHz limitation probably won’t really be an issue though as you can see oscillation as fuzz, even though the scope might not trigger the sweep to see the oscillation correctly. But these scopes are more meant for audio frequencies, like fault-finding audio signal paths through equipment.
Of course if you’re developing circuitry it would be good to have a nice scope and you can get decent S/H ones for reasonable prices. I’d be looking for a digital oscilloscope (DSO) as you can see low frequency signals better. The waveform ‘persists’ on the display at slow speed. You can also get PC scopes that use your computer as the display and for its brain power, and a separate small USB box is what you connect the probes to.(Like PicoScope etc).
For occasional audio-only use (no hunting for oscillations!) you can turn your phone into an oscilloscope as well with the mic input, or with a USB A-D converter.

AMR-iFi R&D

A quick update - the buzz is no longer obviously audible, I don't know why this would be but it seems that I can only detect it now with ear to speaker - any thoughts?

We are what we pretend to be, so we must be careful about what we pretend to be.