(AK) classA HA/Preamplifier.

[jonclancy]

Thanks for indicating a matching target, Alex. Trouble is that these meters resolve to around .3% and that's when the fun starts...

So, 2% is a target = around 4 Hfe or so.

I was wondering about theHfe match between the 5171 and the 1930s as mine are a world apart.

Cheers

Jon

[Deleted]

Mar 2, 2011 20:37:05 GMT jonclancy said:

Thanks for indicating a matching target, Alex. Trouble is that these meters resolve to around .3% and that's when the fun starts...

So, 2% is a target = around 4 Hfe or so.

I was wondering about theHfe match between the 5171 and the 1930s as mine are a world apart.

Cheers

Jon

Jon
The actual reading will vary from meter to meter, and is governed more by the testing base current.
For example, my Electronics Australia designed tester has 2 different base current settings.One is for small signal devices.
The higher setting is designed for medium power devices.
My low range corresponds very well with the other results quoted, but a more realistic figure in this case may be with a higher base current of 100uA , where HFE is likely to drop to around 120 or so.(Beta droop.)
Alex

[Deleted]

My low range corresponds very well with the other results quoted, but a more realistic figure in this case may be with a higher base current of 100uA , where HFE is likely to drop to around 120 or so.(Beta droop.)Alex

I just looked up the spec on my meter, it uses a base current of 10uA.

Like Jon, I was more interested in the match between 5171 and 1930. As Alex said, with a couple of exceptions, my like type are pretty close but across types they fall way outside the 2% parameter.

It's a little academic for me at the mo as I have more Toshes on the way and I have not gone into offsetting low hfe in one place for higher hfe elsewhere.

I really only posted my readings to lend life to the idea.

[XTRProf]

Mar 2, 2011 8:43:56 GMT @sandyk said:

Mar 2, 2011 7:59:41 GMT XTRProf said:

Really just in case if I can't find 0.1uf film caps in Singapore, can I subsitute those for 0.15uf film caps. I remember I can get it from one of the diy shops in Singapore. I know for sure that C4, C6, C9, C10, C18, C19 and C20 can. But I'm not too sure for C3 though.

Thanks for any feedback.

Chong
150nF should be fine at all locations, but 100nF should normally be easier to obtain than 150nF as they are far more commonly used. Alex

Thanks Alex and noted. That made my "job" easier.

[Deleted]

Mar 3, 2011 0:15:42 GMT @cjarchez said:

My low range corresponds very well with the other results quoted, but a more realistic figure in this case may be with a higher base current of 100uA , where HFE is likely to drop to around 120 or so.(Beta droop.)Alex

I just looked up the spec on my meter, it uses a base current of 10uA.

Like Jon, I was more interested in the match between 5171 and 1930. As Alex said, with a couple of exceptions, my like type are pretty close but across types they fall way outside the 2% parameter.

It's a little academic for me at the mo as I have more Toshes on the way and I have not gone into offsetting low hfe in one place for higher hfe elsewhere.

I really only posted my readings to lend life to the idea.

Hi Chris
My tester is also 10uA on the low scale.
If we wanted to be really anal about matching the output devices,
we would also match the PNP and NPN for VBE.
If you try this you are in for a shock, because although they may have similar HFE, and are claimed to be complementary,
you will find due to different fabrication, the VBEs will not normally match !
Regards
Alex

[XTRProf]

Hi guys,

Don't get too "depressed" over not matching exactly. Like for Chris's 5171s, those 189 and 190 are as good as considered matched. I don't think our ears are so sensitive (don't forget most of us are > 40 ) as to be able to hear a difference of 1, 2, 3 or even 4 although the bigger overall picture is every small things add up to become a "mountain". But then don't forget there is the subtractive as well as the additive side of the "errors" of all the circuit components and so it comes back to the same "neutrality" that we are trying our best to aim for. Anyway, what we are aiming for now is just STATIC very close match. When under DYNAMIC condition and when the heat is on, yeah, literally when we power the components up, all matches WILL DRIFT to a certain extent. Up or down, we really don't know for some unless we measure it under dynamic condition. That will be to another level of highend methodology. If we have the time and the right measuring equipment and ways, that can be explored.

Well, anyone wants to graduate to the SUPERNUT level yet? No, definitely not me yet, given the time and equipment that I have now, as I'm happy with my current NUT status.

[jonclancy]

My DCA55 measured Vbe as .69 and .7V on all my 1930 and 5171 - sounds about right...

[Deleted]

Mar 3, 2011 9:14:29 GMT jonclancy said:

My DCA55 measured Vbe as .69 and .7V on all my 1930 and 5171 - sounds about right...

Jon
Actually, it is very wrong!
It's because it uses very high base current to identify different types of semiconductors.
In the Class A HA for example, the voltage across the 100R resistors base to emitter is 595mV or less. Small signal transistors are more typically 620mV or therabouts in most circuits.Try looking at VBE readings in a JLH .
Alex

[XTRProf]

Mar 3, 2011 9:43:13 GMT @sandyk said:

Mar 3, 2011 9:14:29 GMT jonclancy said:

My DCA55 measured Vbe as .69 and .7V on all my 1930 and 5171 - sounds about right...

Jon
Actually, it is very wrong!
It's because it uses very high base current to identify different types of semiconductors.
In the Class A HA for example, the voltage across the 100R resistors base to emitter is 595mV or less. Small signal transistors are more typically 620mV or therabouts in most circuits.Try looking at VBE readings in a JLH .
Alex

It´s very correct. Your reference is 595mV to 620mV, right. Jon´s measurements were 0.69V = 690mV and 0.7V = 700mV.

[Deleted]

Mar 3, 2011 10:07:13 GMT XTRProf said:

Mar 3, 2011 9:43:13 GMT @sandyk said:

Jon
Actually, it is very wrong!
It's because it uses very high base current to identify different types of semiconductors.
In the Class A HA for example, the voltage across the 100R resistors base to emitter is 595mV or less. Small signal transistors are more typically 620mV or therabouts in most circuits.Try looking at VBE readings in a JLH .
Alex

It´s very correct. Your reference is 595mV to 620mV, right. Jon´s measurements were 0.69V = 690mV and 0.7V = 700mV.

Chong
I think we are at cross purposes here.
The readings that Jon is getting are indicative of unusually high base currents due to the tester, and NOT what you would see when using a DMM with transistor test facilities or other types of tester.
They are meaningless to anybody who doesn't use that type of tester.
Alex

[jonclancy]

Ok, I got my 0.7V cross-check from here:

www.ehow.com/how_5487134_read-transistors.html

Para 6.

I know we can go too deeply into this matching malarky, but as I have got a new toy, I want to give it an outing. Plus, I know that I will have built this amp to the best of my (limited) ability.

A more serious question as I try to source inexpensive 79L15:

18V trafo will give around 25V rectified. 22 or 25V will go up to and over 30V. I'm looking at a 50VA 2x18V model at the moment. 50VA is only a couple of quid more expensive than a 30VA model. Any reasons not to?

CHeers

Jon

[Deleted]

A more serious question as I try to source inexpensive 79L15:

Jon,

I had the same problem, unless I wanted a bag full of 'em!

Shaun kindly sent a couple of spares he had. I have since found a good source at Banzai whom I use from time to time and now there seem to more choices on ebay too.

If you only need 2 then I now have spares! Let me know.

Chris

[Deleted]

Mar 3, 2011 16:02:41 GMT jonclancy said:

Ok, I got my 0.7V cross-check from here:

www.ehow.com/how_5487134_read-transistors.html

Para 6.

I know we can go too deeply into this matching malarky, but as I have got a new toy, I want to give it an outing. Plus, I know that I will have built this amp to the best of my (limited) ability.

A more serious question as I try to source inexpensive 79L15:

18V trafo will give around 25V rectified. 22 or 25V will go up to and over 30V. I'm looking at a 50VA 2x18V model at the moment. 50VA is only a couple of quid more expensive than a 30VA model. Any reasons not to?

CHeers

Jon

Hi Jon

i also have 79L15 if you need some. apparently they are out of production according to farnell. i sourced mine from eBay.

PM me if you want some.

take care

[Will]

How bizarre is the 79L15 situation? It's a standard, normal part, and there were loads when I was putting the pcb design together.

It was the same with bc550/560 a while back as well.

Thing is, I reckon that this is a taste of things to come, as more of out taken for granted thru-hole parts become smd only.

[Deleted]

Mar 3, 2011 17:04:07 GMT @shaun said:

Mar 3, 2011 16:02:41 GMT jonclancy said:

Ok, I got my 0.7V cross-check from here:

www.ehow.com/how_5487134_read-transistors.html

Para 6.

I know we can go too deeply into this matching malarky, but as I have got a new toy, I want to give it an outing. Plus, I know that I will have built this amp to the best of my (limited) ability.

A more serious question as I try to source inexpensive 79L15:

18V trafo will give around 25V rectified. 22 or 25V will go up to and over 30V. I'm looking at a 50VA 2x18V model at the moment. 50VA is only a couple of quid more expensive than a 30VA model. Any reasons not to?

CHeers

Jon

Hi Jon

i also have 79L15 if you need some. apparently they are out of production according to farnell. i sourced mine from eBay.

PM me if you want some.

take care

Seems that I better grab a few from Jaycar and Altronics while they still have them !

P.S.
RS appears to have them available from several different manufacturers.
It seems unlikely that they would all cease production of this part.

[jonclancy]

I can get them from both Farnell and RS - but I wanted a bag of cheapies from Tosh at 8p. Not 40p! I can get Taiwan Semiconductor manufactured ones at around 20p each for 20 or so. I suppose they'll all be the same...

My trafos will take me over the Farnell free postage threshold, so might just bite the bullet and go with them.

Cheers

Jon

[Will]

18V trafo will give around 25V rectified. 22 or 25V will go up to and over 30V. I'm looking at a 50VA 2x18V model at the moment. 50VA is only a couple of quid more expensive than a 30VA model. Any reasons not to?

Hi Jon,

No problem with a higher VA rating, just means that you have plenty in reserve TBH, 30VA is overkill, using full wave rectification. Using higher than 18Vac (and developing close to 30Vdc) would also be OK, but fit bigger heatsinks to your regs, as it'll just be given off as waste heat. Mind you, it does bring the possibilty of pre-regulation into play...

[Will]

eBay link for some nxp (philips) bc327-40 transistors, might be useful to some.

cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=180627520066&ssPageName=STRK:MEWNX:IT#ht_500wt_922

Jon, I'll email you, as I think I've found a few things that I need from RS

[Will]

Mar 2, 2011 12:58:47 GMT @cjarchez said:

Re the tosh swaps, I found the correct box!
So I'll kick things off with the balance of my "Audio Wind" pieces.



Note I'm using a standard meter, and had a bit of a game getting a constant reading.

Hi Chris,

Looks a pretty good quad of the 196/190 hfe toshibas, very nice!

When you say it took a while to stabilise, it could be that the transistor had warmed up when you handled them, placing them in the DMM. When doing that, I've left them for a minute or so before testing, or used pliers to put them in the dmm.

[Deleted]

Mar 3, 2011 21:51:02 GMT Will said:

Hi Chris,

Looks a pretty good quad of the 196/190 hfe toshibas, very nice!

When you say it took a while to stabilise, it could be that the transistor had warmed up when you handled them, placing them in the DMM. When doing that, I've left them for a minute or so before testing, or used pliers to put them in the dmm.

Exactly that Will. They were very cold which was causing the problem, the reading rising before my eyes.
I proceeded to warm each one for a few seconds between finger and thumb. Then left them to return to a normal room temperature.

I then managed a stable reading

[XTRProf]

Mar 3, 2011 16:02:41 GMT jonclancy said:

I'm looking at a 50VA 2x18V model at the moment. 50VA is only a couple of quid more expensive than a 30VA model. Any reasons not to?

Go for the 50VA. For me, I will go SUPERNUT overboard for this moment at >100VA. It will sound better. But why not try R-Core or C-Core traffos for a different flavour like what I did. I actually wanted >100VA R-Core traffos in my implementation but can't get locally.

[XTRProf]

Mar 3, 2011 19:48:45 GMT Will said:

Thing is, I reckon that this is a taste of things to come, as more of out taken for granted thru-hole parts become smd only.

That's still not the worst part. Wait till everything is IC ...........

[Deleted]

I will go SUPERNUT overboard for this moment at >100VA. It will sound better

Why ?
I tried a 50VA toroidal with my X-DAC V3 and it sounded worse.
Tighter coupling between windings may cause a further reduction in
EMI rejection ?
It's a good question for Frans !

[jonclancy]

A good question, Alex.

The space for mounting the trafos in my case will be a consideration as well.

[jeffc]

Hi Guys,

I’m home after a couple of weeks off the beaten track in India and Vietnam sampling prawns, home a day and a half and then off again for the best part of 2 days driving 1800 km with my son #2 to Melbourne, flight home late Monday, work on Tuesday. Still rocking and rolling, I got a pleasant surprise Tuesday arvo with my boards arriving home from their holiday in Sydney with Alex to see what might be amiss with one. Good news from Alex is that both now appear to be performing to spec with the offset correctors doing their job 0.00000 mV offset. Minor exaggeration but close to zero. Plus Will’s gift for PCB design has Alex’s amp sounding maybe even nicer than his old work horse. Might have been teasing here but he seems pretty contented. So all good for me to get this amp cased up and working.

If interested, I used Panasonic NHG caps (not so low ESR) and a couple of Nichicon FG caps for the offset corrector because a friend had surplus to requirements. For the input 220R and output 68R resistors I used to RPRs that I had some spare and spare OPA234 opamps were used for the offset corrector. Good to have spares. In all values availabe, I used the new copper-leaded MF resistors Jaycar are now moving to. Nice touch. Alex also chipped in kindly with 3V3 zenners and some film caps for the Class A PCB as well as JLHs, thanks mate.

Pic of one of the boards (note: link now fitted across gound lift)


Pic of potential board and transformer layout placed on the lid on my case


Seeing there is some paranoia about AC and transformers in the same case as the Class A HA, JLH and PSU boards, you’ve got me paranoid now. So can you see any general problems with this layout? If transformers are placed at the rear of the case, I might use an on/off switch at the rear left so that AC is contained at that end and run shielded signal cables from right-side RCAs along that side to the stepped potentiometer on the right, headphone jack placed centrally at the front.

cheers..jeffc