Hey everyone, I thought you might find this an interesting little story.
I am in the middle of refurbishing and, hopefully, repairing my old XL280 and have uncovered some mysteries concerning this amp's past. I bought this amp as a showroom model that had been 'repaired'. As a 19 year old kid, I never thought to check under the hood - it sounded great.
Well, now that it's apart I can tell you that this poor thing was used as some kind of 'junker for parts' before it was repaired and sold to me.
Most of the caps do not match the original spec parts except in capacitance, and the wiring was left a little sloppy. Resistors appear original (even most carbon films are within .1-.5% tolerance - right on Hafler!) and output MOSFETs are original.
By far, the biggest oddity of all was when I found 4 pairs of NEC K163/J44 JFETs in place of the original K147/J72 pairs. I don't know if the K147/J72's were stolen or blown up.
I have been planning on replacing the K163/J44 pairs with K170BL/J74BL's but I just read that the K163/J44 pairs were used in Krell amps and are considered to be high quality parts.
If it was your project, what would you do?
The amp has functioned for 17 years with these K163/J44 pairs, should I even bother to change them? I have absolutely never read anything about anything but K147/J72 or K170/J74 pairs in this amp, and I would like it to sound as 'good' as possible of course.
If I did pull these K163/J44's, I would be willing to send them to a member of this forum, free of charge, if they need them.
I am in the middle of refurbishing and, hopefully, repairing my old XL280 and have uncovered some mysteries concerning this amp's past. I bought this amp as a showroom model that had been 'repaired'. As a 19 year old kid, I never thought to check under the hood - it sounded great.
Well, now that it's apart I can tell you that this poor thing was used as some kind of 'junker for parts' before it was repaired and sold to me.
Most of the caps do not match the original spec parts except in capacitance, and the wiring was left a little sloppy. Resistors appear original (even most carbon films are within .1-.5% tolerance - right on Hafler!) and output MOSFETs are original.
By far, the biggest oddity of all was when I found 4 pairs of NEC K163/J44 JFETs in place of the original K147/J72 pairs. I don't know if the K147/J72's were stolen or blown up.
I have been planning on replacing the K163/J44 pairs with K170BL/J74BL's but I just read that the K163/J44 pairs were used in Krell amps and are considered to be high quality parts.
If it was your project, what would you do?
The amp has functioned for 17 years with these K163/J44 pairs, should I even bother to change them? I have absolutely never read anything about anything but K147/J72 or K170/J74 pairs in this amp, and I would like it to sound as 'good' as possible of course.
If I did pull these K163/J44's, I would be willing to send them to a member of this forum, free of charge, if they need them.
i'm probably speaking for myself as well as others ... if the jfets currently in circuit now are working, leave 'em alone. 
you will get bigger changes (improvements???) in sound by doing other things, like increasing output stage bias, etc.
you bring up a good point though. with the scarcity of toshiba complementary jfets continuing to get worse, how long before old gear gets snarfed up only to be used as a source for expensive jfets ...
mlloyd1
btw, forgot the link to the datasheet for the 2sj44 in case you were curious: http://www.necel.com/nesdis/image/TC-3344A.pdf
you will get bigger changes (improvements???) in sound by doing other things, like increasing output stage bias, etc.
you bring up a good point though. with the scarcity of toshiba complementary jfets continuing to get worse, how long before old gear gets snarfed up only to be used as a source for expensive jfets ...
mlloyd1
btw, forgot the link to the datasheet for the 2sj44 in case you were curious: http://www.necel.com/nesdis/image/TC-3344A.pdf
Many thanks
Thank you mlloyd1,
I knew you would probably have some info for me. It's great to know there is another XL-280 fan out there. It says something that I could love this amp this much after all these years with compromised guts.
As far as the 'bad side' saga goes, well...
I have that side in pieces again, resistance across D-S for Q1,Q2,Q3, and Q4 (the JFETs) is:
Q1=56 Ohms, Q2=76 Ohms, Q3=50 Ohms, and Q4=38 Ohms.
Sound Ok?
I have already ordered 20 and 20 of the 2N5550/2N5401 pairs from ON Semi in case the cascodes are an issue. I upgraded the 2N5550's to 2N5551's. Tougher and less noise to boot.
I have noticed that Hafler may have had a typo in his schematics concerning the 'NP2222A'. Everything I can find points to this actually being a 'PN2222A'.
I have been able to source some caps that should far outperform what was in here. For example, I found a 400V .01uF Metallized Polyester at C4. Terrible. Spec is a 160V PP.
Oh well, that's all for now, gotta go - maybe I'll post pics later.
Thank you mlloyd1,
I knew you would probably have some info for me. It's great to know there is another XL-280 fan out there. It says something that I could love this amp this much after all these years with compromised guts.
As far as the 'bad side' saga goes, well...
I have that side in pieces again, resistance across D-S for Q1,Q2,Q3, and Q4 (the JFETs) is:
Q1=56 Ohms, Q2=76 Ohms, Q3=50 Ohms, and Q4=38 Ohms.
Sound Ok?
I have already ordered 20 and 20 of the 2N5550/2N5401 pairs from ON Semi in case the cascodes are an issue. I upgraded the 2N5550's to 2N5551's. Tougher and less noise to boot.
I have noticed that Hafler may have had a typo in his schematics concerning the 'NP2222A'. Everything I can find points to this actually being a 'PN2222A'.
I have been able to source some caps that should far outperform what was in here. For example, I found a 400V .01uF Metallized Polyester at C4. Terrible. Spec is a 160V PP.
Oh well, that's all for now, gotta go - maybe I'll post pics later.
Ahh yes... The NP2222A is actually a PN2222A. IIRC thats just the transistor in the biasing circuit and can be replaced with any decent npn with a high hfe. I used a 2n3904 in place of the 2222A in one of my amps without any issues.
Also, Hafler was known to use what parts they could find to meet price demands. I have found several mismatched parts in my amps. Usually its only capacitors and resistors, parts that do not make much of a differance to the operation of the amp. It is odd they would use a differant jfet, but it is possible.
Also, Hafler was known to use what parts they could find to meet price demands. I have found several mismatched parts in my amps. Usually its only capacitors and resistors, parts that do not make much of a differance to the operation of the amp. It is odd they would use a differant jfet, but it is possible.
mlloyd1 said:
M LLoyd is right. The NEC's were excellent transistors, and very comparable to the 2SK170/2SJ74 pairs. I wish they were still available, as I would be using them.
Both of the above devices have about half the transconductance of the original 2SK147/2SJ72 parts. So you may have to change the value of the source resistors to optimize things fully. But if the amp is working well, I would leave it alone.
sorry about that, Mloyd - I unnerstand now. NEC was a pretty low-key and underregarded manufacturer of small signal stuff back in the 80's. I have a bag of 2SJ45s from NEC that has been languishing for lack of an application. It would be a one-time thing, though, as I have no idea where I'd get replacements.
I am assuming R1, R2, and R3 are the source resistors... Correct me if I am wrong.
R1=2210 Ohms, R2=47,500, and R3=110,000
All the resistors in this amp test remarkably close spec, including R1, R2, and R3. There are no 'mods' to the values of anything except the JFETs and capacitor voltage ratings.
I have some Dale RN60D 47.5K and 110K that test low (like 47.2K and 109K) - maybe worth a try while it's apart?
I must say, I consider myself an Ok technician, but I am not qualified to re-engineer this baby (nor do I want to), I would, however, throw in a couple off value premium resistors if it would likely improve my sound.
I will wait to hear what the forum has to say on this.
I dragged out my tattered old NEC data book and found out that the 2SJ44/2SK163 have breakdown voltages of 40/50V respectively, noise figures of ~1.5 nV/root HZ, and transconductance of around 9 mS. Not bad at all, and the relatively high breaksdown voltage of the p-channel device is attractive (a 2SJ74 is only 25V). Pinout is the same as 2SJ74/2SK170.
Since the XL-280 cascodes the input differential pairs, one could conceivably sub the Toshiba 2SJ74/2SK170 (matched pairs, of course). The characteristics are similar to the canned matched pairs originally used, and you could select for tighter matching than guaranteed by the 2SJ73/2SK146 specs. The impetus for doing this would be ~2X higher gain in the input differential stage, resulting in higher open loop gain and perhaps a bit lower distortion. The Toshiba devices also have a lower noise figure than the NECs, but that doesn't matter too much for a power amplifier.
Since the XL-280 cascodes the input differential pairs, one could conceivably sub the Toshiba 2SJ74/2SK170 (matched pairs, of course). The characteristics are similar to the canned matched pairs originally used, and you could select for tighter matching than guaranteed by the 2SJ73/2SK146 specs. The impetus for doing this would be ~2X higher gain in the input differential stage, resulting in higher open loop gain and perhaps a bit lower distortion. The Toshiba devices also have a lower noise figure than the NECs, but that doesn't matter too much for a power amplifier.
Hi wrenchone,
Thanks for the specs. If I had the money, I would order 4 pairs of 2SK147/2SJ72 JFETs from Erno Borbely right this second - but it's roughly $150. I have heard that same advice regarding the lower noise parts not being necessary for a power amp, but... It's lower noise, you know what I mean
!
After all, my DH-110 preamp does have a 2SK147 in the line amp.
For now, I will leave the NEC's alone, especially considering Mlloyd's suggestion that they may even be original parts.
I am not planning on deviating from the schematic, but of course I am naturally curious about the comment made by Charles Hansen. I attached the schematic of the cascodes/JFETs.
R11, 12, 13, and 14 are 47 Ohms.
R15 in the middle is a 68 Ohm.
In THEORY, what would optimize this circuit for less transconductance?
Thanks for the specs. If I had the money, I would order 4 pairs of 2SK147/2SJ72 JFETs from Erno Borbely right this second - but it's roughly $150. I have heard that same advice regarding the lower noise parts not being necessary for a power amp, but... It's lower noise, you know what I mean
!After all, my DH-110 preamp does have a 2SK147 in the line amp.
For now, I will leave the NEC's alone, especially considering Mlloyd's suggestion that they may even be original parts.
I am not planning on deviating from the schematic, but of course I am naturally curious about the comment made by Charles Hansen. I attached the schematic of the cascodes/JFETs.
R11, 12, 13, and 14 are 47 Ohms.
R15 in the middle is a 68 Ohm.
In THEORY, what would optimize this circuit for less transconductance?
An externally hosted image should be here but it was not working when we last tested it.
HaflerFreak said:
Just leave it be. As long as it is working, count your blessings.
If you really, really want it to be original, you will have to buy the original transistors from Erno Borbely. He is the only person I know that is still selling them.
The next closest thing would be to purchase about 20 each of the Toshiba K170 and J74 pairs and then match them. Paralleling two matched K170's will be virtually identical to a K147, and similarly for the P-channel parts.
But to try to change the circuit for the NEC transistors will be an exercise in futility. Reducing the source resistors will increase the transconductance, but also make the matching more critical. (You can get a good idea of how well the existing transistors are matched by measuring the DC voltage across each source resistor while the amp is "on" and has no signal.)
But then you would need to change the middle resistor to adjust the overall current levels. One problem is that you don't know what the original current was. The Toshiba parts come in three different grades, and the schematic doesn't specify which grade was used. The NEC's come in four different grades, and if you don't have the right grade you won't be able to reach the proper current level anyway.
Finally, you would want to change the gate resistors also to optimize them for the lower value of input capacitance that the NEC parts have.
And when all that is done, what will you have accomplished? Will the amp sound better? I doubt it. Will it even sound different? I doubt it.
If only I could afford college
Thank you Charles,
Believe it or not, that is almost the response that I suspected. I am not trying to tweak or perfect this amplifier, the reason for my posts is a still (after 2 months) unidentified failure that appears to be in the right channel of this amp.
I don't have a signal generator, or an oscilloscope, but I do have a little Variac, 3 DMM's, a static strap, a precision cap tester, and my ears/common sense.
I suspect it may be input related because the 'failure' occurred at a fairly low listening level under odd input conditions.
I was feeding the amp with a preamp with a variable out from a TV that was receiving its source from a DVD - say that ten times.
There was some distortion, not overly loud, but if 'overload' has a sound, this was it. I guess if you could describe the sound of a capacitor giving way, this might be it.
The problem is, I have dissected this thing twice and no burnt parts - no burnt resistors, no burnt capacitors - everything tests fine. Supply voltages test within range of operation. Idle current is fairly stable and is, at the moment, adjusted a little low.
Just about the only thing I have not tested at this point is the transistors.
Oh yeah, I almost forgot.
I didn't measure the current at the sources of Q1,Q2,Q3,Q4 when I had this thing together, but I measured the voltages. As far as matching goes:
Q1 (K163) Source voltage at idle: 316 mV
Q2 (K163) Source voltage at idle: 340mV
Q3 (J44) Source voltage at idle: 269mV
Q4 (J44) Source voltage at idle: 290mV
Problem is, my bass is gone in the right channel. Actually, it seems to be there - but cut off or attenuated. The entire channel seems to have less 'gain' since the blowout.
The offset measures about -14 mV in that channel as opposed to -4mV on the good channel.
The amp kinda sorta functions - but not really.
It mystifies me, still, I trust myself and this forum to find the answer as opposed to some cocky, over-priced, supposedly educated tech.
Thank you Charles,
Believe it or not, that is almost the response that I suspected. I am not trying to tweak or perfect this amplifier, the reason for my posts is a still (after 2 months) unidentified failure that appears to be in the right channel of this amp.
I don't have a signal generator, or an oscilloscope, but I do have a little Variac, 3 DMM's, a static strap, a precision cap tester, and my ears/common sense.
I suspect it may be input related because the 'failure' occurred at a fairly low listening level under odd input conditions.
I was feeding the amp with a preamp with a variable out from a TV that was receiving its source from a DVD - say that ten times.
There was some distortion, not overly loud, but if 'overload' has a sound, this was it. I guess if you could describe the sound of a capacitor giving way, this might be it.
The problem is, I have dissected this thing twice and no burnt parts - no burnt resistors, no burnt capacitors - everything tests fine. Supply voltages test within range of operation. Idle current is fairly stable and is, at the moment, adjusted a little low.
Just about the only thing I have not tested at this point is the transistors.
Oh yeah, I almost forgot.
I didn't measure the current at the sources of Q1,Q2,Q3,Q4 when I had this thing together, but I measured the voltages. As far as matching goes:
Q1 (K163) Source voltage at idle: 316 mV
Q2 (K163) Source voltage at idle: 340mV
Q3 (J44) Source voltage at idle: 269mV
Q4 (J44) Source voltage at idle: 290mV
Problem is, my bass is gone in the right channel. Actually, it seems to be there - but cut off or attenuated. The entire channel seems to have less 'gain' since the blowout.
The offset measures about -14 mV in that channel as opposed to -4mV on the good channel.
The amp kinda sorta functions - but not really.
It mystifies me, still, I trust myself and this forum to find the answer as opposed to some cocky, over-priced, supposedly educated tech.
What I really mean
Hey everyone,
You know, after reading my last post I thought I should probably clarify my position regarding comments I have made about repair techs.
First off, I am very thankful for all of the information and helpful comments that I have received from this forum, especially MLloyd1. This website is a God send, and I have every intention of making a donation to DIYAudio as soon as I have the means. Heck, if I can, I will be taking my amp to a shop when I can to have it put on a scope and checked.
I have noticed that there are experts here that sometimes give advice that is quite literally worth money - and they do it out of the kindness of their hearts. I have NEVER meant any disrespect towards the electrical engineers, experienced repair professionals, or genius hobbyists.
Any disparaging remarks that I have made are in reference to cocky, dishonest people that are well paid to perform a job that they are not qualified to do. Unfortunately, these kinds of people are showing up as auto mechanics, teachers, politicians, managers, and home theater installers.
It is pathetic.
I am an ignorant, undereducated audiophile who had his heyday 17 years ago. I have tested with a 130+ IQ and been asked all my life, 'Why don't you become a doctor?'.
I am a pizzaman.
I will always be a pizzaman (and not an EE like I dream of) because I have used all my savings and my average tip has been going down steadily every year for the past 4 years. I make so little that I literally cannot stop in order to escape.
I feel bad when I ask stupid questions, because I know you pros are wondering why I would even bother trying to fix my own stuff, but I have always been an ace with a soldering iron and I think I know the electrical basics from studying to be a computer technician - I was discharging TV flybacks when I was 10!
So, anyhow, I just wanted the REAL brainiacs to know that I greatly admire you, and I am just bitter from being burned so many times by stupid, arrogant fools.
Have a good night all, on to my project....
Hey everyone,
You know, after reading my last post I thought I should probably clarify my position regarding comments I have made about repair techs.
First off, I am very thankful for all of the information and helpful comments that I have received from this forum, especially MLloyd1. This website is a God send, and I have every intention of making a donation to DIYAudio as soon as I have the means. Heck, if I can, I will be taking my amp to a shop when I can to have it put on a scope and checked.
I have noticed that there are experts here that sometimes give advice that is quite literally worth money - and they do it out of the kindness of their hearts. I have NEVER meant any disrespect towards the electrical engineers, experienced repair professionals, or genius hobbyists.
Any disparaging remarks that I have made are in reference to cocky, dishonest people that are well paid to perform a job that they are not qualified to do. Unfortunately, these kinds of people are showing up as auto mechanics, teachers, politicians, managers, and home theater installers.
It is pathetic.
I am an ignorant, undereducated audiophile who had his heyday 17 years ago. I have tested with a 130+ IQ and been asked all my life, 'Why don't you become a doctor?'.
I am a pizzaman.
I will always be a pizzaman (and not an EE like I dream of) because I have used all my savings and my average tip has been going down steadily every year for the past 4 years. I make so little that I literally cannot stop in order to escape.
I feel bad when I ask stupid questions, because I know you pros are wondering why I would even bother trying to fix my own stuff, but I have always been an ace with a soldering iron and I think I know the electrical basics from studying to be a computer technician - I was discharging TV flybacks when I was 10!
So, anyhow, I just wanted the REAL brainiacs to know that I greatly admire you, and I am just bitter from being burned so many times by stupid, arrogant fools.
Have a good night all, on to my project....
VintageAmp -
Hopefully not doomed, but I must admit I really never see these FETs for sale on the trading post. I have not checked eBay, but it might be a place to keep an eye on.
Suggestion: I would try to find out what commercial amps used these FETs (I'm sure people here could help) and hunt down the cheapest one you can find (probably on eBay) then cannabilize it for the FETs.
Also, do you need these specific FETs for their high voltage capabilities or would 2SK170/2SJ74 do?
I found two 2SK163 FETs used in my DH110 preamp for muting FETs, but these would most likely not be matched and would still leave you looking for J44's. Still, an old broken 110 could prove useful.
Oh yeah, look for XL-280's for sale and inquire about their FETs, surely mine is not the only one out there with k163/j44.
Good luck!
Hopefully not doomed, but I must admit I really never see these FETs for sale on the trading post. I have not checked eBay, but it might be a place to keep an eye on.
Suggestion: I would try to find out what commercial amps used these FETs (I'm sure people here could help) and hunt down the cheapest one you can find (probably on eBay) then cannabilize it for the FETs.
Also, do you need these specific FETs for their high voltage capabilities or would 2SK170/2SJ74 do?
I found two 2SK163 FETs used in my DH110 preamp for muting FETs, but these would most likely not be matched and would still leave you looking for J44's. Still, an old broken 110 could prove useful.
Oh yeah, look for XL-280's for sale and inquire about their FETs, surely mine is not the only one out there with k163/j44.
Good luck!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.