Bob, agreed.
I could not match the NPNs to the PNPs.
So I matched the NPNs to each other and the PNPs to each other.
The consequence is that the halves of the LTP track each other for temperature and thus minimise the change in stage balance with variations in temperature.
The other consequence of not matching the hFE of the PNPs to the NPNs is increased input offset current. This gives rise to input offset voltage. But does not affect setting up output offset.
However, if the input offset current changes with temperature then the output offset voltage will also vary.
The solution is to thermally couple all the pairs and to keep them cool to minimise changes and /or differences in temperature.
I could not match the NPNs to the PNPs.
So I matched the NPNs to each other and the PNPs to each other.
The consequence is that the halves of the LTP track each other for temperature and thus minimise the change in stage balance with variations in temperature.
The other consequence of not matching the hFE of the PNPs to the NPNs is increased input offset current. This gives rise to input offset voltage. But does not affect setting up output offset.
However, if the input offset current changes with temperature then the output offset voltage will also vary.
The solution is to thermally couple all the pairs and to keep them cool to minimise changes and /or differences in temperature.
Thanks for the feedback guys. As Bob suggests, I'm trying to match to .001 - well actually my meter doesn't have that resolution. But I did take the matching recommendation literally (noob)
Tad's offset results are encouraging. Maybe a bit lucky too. But as Andrew mentions, rise in input offset voltage won't affect output offset. I have dozens of P's matched to each other, same for N's. Tad, no effort to match small signal or outputs??
The lucky group of 8 will go into the tweeter amp
One thing I found was that the Onsemi's has superior gain numbers over the Fairchilds for both types. My manufacturer mix -n- match strategy worked pretty good because of gain difference of N & P types.
Tad's offset results are encouraging. Maybe a bit lucky too. But as Andrew mentions, rise in input offset voltage won't affect output offset. I have dozens of P's matched to each other, same for N's. Tad, no effort to match small signal or outputs??
The lucky group of 8 will go into the tweeter amp
One thing I found was that the Onsemi's has superior gain numbers over the Fairchilds for both types. My manufacturer mix -n- match strategy worked pretty good because of gain difference of N & P types.
Driver sinking & "boutique" caps
I hope my frequent questions don't get me ostracized, but...
a)What is the best way to mount the 15032/15033 driver transistors?
I see the pix on Jens' website, but am not sure exactly what I'm seeing. It looks like some type of spacer or standoff is used under the drivers(is this made of aluminum?).
I have some TO220 heat sinks here, but notice they're not used...
b)If using ultra low ESR electrolytics(like MuseKZ or Black Gate
) at C1,C3,C10, & C14, is there still need (or could it degrade sound? 
) to use the specified 1uf bypasses (it's hard to find premium 1uf to fit the space allotted)? I've decided to go all-out in choosing parts quality to coax the most from these amps.
No flame war please!
Thanks in advance, as usual
-Chas
I hope my frequent questions don't get me ostracized, but...
a)What is the best way to mount the 15032/15033 driver transistors?
I see the pix on Jens' website, but am not sure exactly what I'm seeing. It looks like some type of spacer or standoff is used under the drivers(is this made of aluminum?).
I have some TO220 heat sinks here, but notice they're not used...
b)If using ultra low ESR electrolytics(like MuseKZ or Black Gate
) at C1,C3,C10, & C14, is there still need (or could it degrade sound? ) to use the specified 1uf bypasses (it's hard to find premium 1uf to fit the space allotted)? I've decided to go all-out in choosing parts quality to coax the most from these amps.No flame war please!
Thanks in advance, as usual
-Chas
Boards....
I've had problems soldering the my 330 Ohm resistors. Some of the problems I have had : They take an extraordinary amount of heat to stick; Some take so much heat that they burn the pad; The leads are too large; They refuse to "wick" the solder; I had to solder some of the resistors from the bottom to get TO them and I do not believe the solder has "wicked" to the upper trace on the opposite side of the board. This consequently has put these boards in jeopardy. I tried pulling one out of the easier resistors above the output transistors and it just makes a mess and lifts the trace with it because the leads are so thick.
I have been checking every joint by magnifying glass and cleaning each solder connection with a brass brush to get to get the rosin out of the way. I've found that the thinner leads do a better job on the capillary action and perform the "necessary "wicking" when the hole has some clearance.
What is your take on this? I think most the the parts could be salvaged but the boards may turn out to be junk. I tried reaming out some holes and it does seem to help but at this point I'm kind of skeptical about these boards being that there is no way to check them. Is the correct proceedure to ream the holes if the leads are snug or tight? Is it always a MUST DO to solder from the topside on this type of board?
I've had problems soldering the my 330 Ohm resistors. Some of the problems I have had : They take an extraordinary amount of heat to stick; Some take so much heat that they burn the pad; The leads are too large; They refuse to "wick" the solder; I had to solder some of the resistors from the bottom to get TO them and I do not believe the solder has "wicked" to the upper trace on the opposite side of the board. This consequently has put these boards in jeopardy. I tried pulling one out of the easier resistors above the output transistors and it just makes a mess and lifts the trace with it because the leads are so thick.
I have been checking every joint by magnifying glass and cleaning each solder connection with a brass brush to get to get the rosin out of the way. I've found that the thinner leads do a better job on the capillary action and perform the "necessary "wicking" when the hole has some clearance.
What is your take on this? I think most the the parts could be salvaged but the boards may turn out to be junk. I tried reaming out some holes and it does seem to help but at this point I'm kind of skeptical about these boards being that there is no way to check them. Is the correct proceedure to ream the holes if the leads are snug or tight? Is it always a MUST DO to solder from the topside on this type of board?
Hi,
it looks like you have iron or steel lead out wires on your 330r resistors.
These should be plated to allow soldering.
If the plating has been removed then they are almost impossible to solder with electrical flux.
You need to pre-tin them, probably with Bakers' fluid, but that's just a guess.
Throw the 330r away and buy new ones with the leadouts correctly tinned and of the correct diameter.
If you do have to ream out PTH holes, then you MUST solder both sides of the PCB to maintain the circuit to the live side trace, it could be both sides are used for traces.
it looks like you have iron or steel lead out wires on your 330r resistors.
These should be plated to allow soldering.
If the plating has been removed then they are almost impossible to solder with electrical flux.
You need to pre-tin them, probably with Bakers' fluid, but that's just a guess.
Throw the 330r away and buy new ones with the leadouts correctly tinned and of the correct diameter.
If you do have to ream out PTH holes, then you MUST solder both sides of the PCB to maintain the circuit to the live side trace, it could be both sides are used for traces.
Brutepuppy,
I posted a caution earlier on in this thread concerning resistors with nice fat leads. The holes in the pcb were spec for a particular size resistor which Jens kindly published on his website. Many more dimensions are found on the Leach 6 transistor page. I had to sand down some of my leads on the Dale Rn resistors and re-tin with solder before using -- These were pure copper leads. It is also very important to solder in the most difficult items first. In my case it was the little 3 legged TO-92 transistors. Everything is done using a mounted well lit magnifying glass for inspection.
One suggestion would be to cut the leads off on both sides of the board and try to solder in new resistors where you have incurred problems top and bottom. It may eliminate the trace lifting. It is not a preferred method, but it has worked for me. Also, try cleaning the board with an approved electronics spray and see if it helps. I ran into this same problem with the solder not sticking on my P 1.7 boards and had to etch them with flux.
Like Andrew said on this amp most traces are used on both sides.
Good Luck Tad
I posted a caution earlier on in this thread concerning resistors with nice fat leads. The holes in the pcb were spec for a particular size resistor which Jens kindly published on his website. Many more dimensions are found on the Leach 6 transistor page. I had to sand down some of my leads on the Dale Rn resistors and re-tin with solder before using -- These were pure copper leads. It is also very important to solder in the most difficult items first. In my case it was the little 3 legged TO-92 transistors. Everything is done using a mounted well lit magnifying glass for inspection.
One suggestion would be to cut the leads off on both sides of the board and try to solder in new resistors where you have incurred problems top and bottom. It may eliminate the trace lifting. It is not a preferred method, but it has worked for me. Also, try cleaning the board with an approved electronics spray and see if it helps. I ran into this same problem with the solder not sticking on my P 1.7 boards and had to etch them with flux.
Like Andrew said on this amp most traces are used on both sides.
Good Luck Tad
Resistors
I have the same resistors as BrutePuppy. Sorry, BP!
Let me know how many resistors you need and I'll try to replace them for you at no cost.
I purchased a load of these at a surplus store for our small group; however, WJH and I have been soldering our boards without any serious problems(the heavy copper traces DO suck up the heat and make soldering a challenge!).
The way I got mine to work is: I heat the pads/leads next to the resistor body(on top of the PCB) until the solder begins to flow then remove the iron immediately, flip the board and solder the other side asap.
I do this one lead at a time, and it is pretty time-consuming. The solder does flow ok(I'm using Wellborne silver-bearing solder and a Weller temp-controlled iron with 800 degree tip). No pads lifted so far*
On my second set of boards, I'll be using Holco, Roederstein and Dales all of which have smaller leads(and cost 3-4x as much as the surplus ones).
And speaking of resistors-
R1/R23 are specified as 100ohms, but I have a bunch of 91ohm/1Watts I was going to use. This is 9% lower than spec-
any forseeable problems using these?
-Chas
*as long as the pads are not connected to a trace (i.e. circular)they shouldn't cause any problems, should they?
I have the same resistors as BrutePuppy. Sorry, BP!
Let me know how many resistors you need and I'll try to replace them for you at no cost.
I purchased a load of these at a surplus store for our small group; however, WJH and I have been soldering our boards without any serious problems(the heavy copper traces DO suck up the heat and make soldering a challenge!).
The way I got mine to work is: I heat the pads/leads next to the resistor body(on top of the PCB) until the solder begins to flow then remove the iron immediately, flip the board and solder the other side asap.
I do this one lead at a time, and it is pretty time-consuming. The solder does flow ok(I'm using Wellborne silver-bearing solder and a Weller temp-controlled iron with 800 degree tip). No pads lifted so far*
On my second set of boards, I'll be using Holco, Roederstein and Dales all of which have smaller leads(and cost 3-4x as much as the surplus ones).
And speaking of resistors-
R1/R23 are specified as 100ohms, but I have a bunch of 91ohm/1Watts I was going to use. This is 9% lower than spec-
any forseeable problems using these?
-Chas
*as long as the pads are not connected to a trace (i.e. circular)they shouldn't cause any problems, should they?
BP - what are you using to solder with? A low power iron and small tip make soldering this board very hard. I have been using a 40W station from http://www.cir.com/ ($40) successfully. Get a variety of tips in multiples. They last a while, but it's reassuring to have a spare and saves shipping costs
When soldering to heavy traces (all on this heavy Cu board) use the largest tip that fits. If you are burning traces, you probably need to go hotter - counter intuitive, but it means that you will get the job done.
HTH
When soldering to heavy traces (all on this heavy Cu board) use the largest tip that fits. If you are burning traces, you probably need to go hotter - counter intuitive, but it means that you will get the job done.
HTH
My non-magnetic Dales were ordered online from Surplussales back east. I guess you could say that they are "NOS", because the package said "February, 1957!" 
I only have 91ohm/1Watts on hand(9% less than spec)-
Since these appear to set the voltage/current to the VAS, is there any forseeable problem using them?
(Bob, Tad, Jens, anyone?)
-Chas
I only have 91ohm/1Watts on hand(9% less than spec)-
Since these appear to set the voltage/current to the VAS, is there any forseeable problem using them?
(Bob, Tad, Jens, anyone?)
-Chas
The basic match is accomplished fairly easily by scaling Prof. Leach's input device matching scheme to your target idle current. Just change the load resistor to an appropriate value. You also need to adjust the base resistor since output devices have much lower current gain than the inputs. Calculate the current gain by comparing base current to output current.
Yes, output matching is supposed to improve sound quality, but my Leach amps sound great without. Vertical MOSFETS need matching to ensure reasonable current sharing.
Chances of getting the Ns to match the Ps are slim.
Yes, output matching is supposed to improve sound quality, but my Leach amps sound great without. Vertical MOSFETS need matching to ensure reasonable current sharing.
Chances of getting the Ns to match the Ps are slim.
It sounds like a few of you are enjoying your little projects. The end result is real nice!!!
If anyone was paying attention I do not believe you will see any really audiophile type parts on Jens boards. He assembled quite a few and had very good results. Just standard quality parts and attention to detail. Dr. Leach in his assembly document specs 5% parts and mostly run of the mill or even cheaper. I think the real quality is in the design and layout.
When I was researching this project I got many nods for the Leach build from some of the most respected members of our forum. Do not sweat the minutae just take your time. Nothing is wrong with soldering in one little part at a time and closely checking it as you go. Remember the chassis will take about 10 times as long to build.
It is nice during the shut in winter months to have a project at hand.
Tad
If anyone was paying attention I do not believe you will see any really audiophile type parts on Jens boards. He assembled quite a few and had very good results. Just standard quality parts and attention to detail. Dr. Leach in his assembly document specs 5% parts and mostly run of the mill or even cheaper. I think the real quality is in the design and layout.
When I was researching this project I got many nods for the Leach build from some of the most respected members of our forum. Do not sweat the minutae just take your time. Nothing is wrong with soldering in one little part at a time and closely checking it as you go. Remember the chassis will take about 10 times as long to build.
It is nice during the shut in winter months to have a project at hand.
Tad
Re: Boards....
BP - I've resorted to using a no-clean low solids liquid flux for improved flow. Like Chas, I'm using an 800º tip to get the heat in quickly.
I have touched up some pads on top, where i didn't flux.
I am just days away from testing the front end of my first PCB. Heat sinks due within the week (I WONT hold my breath though).
Matching zeners was as bad as small signal transistors. I'm beginning to think that 5% is useless! My reference voltages range between 19.2 and 19.8V Maybe I'm anal but I was hoping to exact match 6 boards!
brutepuppy said:
BP - I've resorted to using a no-clean low solids liquid flux for improved flow. Like Chas, I'm using an 800º tip to get the heat in quickly.
I have touched up some pads on top, where i didn't flux.
I am just days away from testing the front end of my first PCB. Heat sinks due within the week (I WONT hold my breath though).
Matching zeners was as bad as small signal transistors. I'm beginning to think that 5% is useless! My reference voltages range between 19.2 and 19.8V Maybe I'm anal but I was hoping to exact match 6 boards!
I loved that thing!