Hi everyone, I just wanted to chime in with my two cents worth and tell ya about the excellent results that I have had replacing the outputs in this fine amp. For anyone having trouble sourcing parts it is possible to replace the outputs with IRFP340PBF and IRFP9240PBF these MOSFET's are inexpensive, readily available from Newark Element 14 and have higher general ratings than the originals and will be more robust. There is an issue with the "bias spreader circuit" that makes it overcompensate with these devices. There is a solution. Some modifications are required but not hard if you have some patience and electronics ability you can perform these mods with great success and the bias circuit will track much better and will be more stable.
You must add a second transistor wired as a thermal diode that gets attached to the heatsink in the stock location. It is wired in series with the original trans and effectively cuts the bias compensation by 50%. The existing trans is relocated to a bracket under the Main PCB which now only compensates for ambient temperature. The modified circuit now tracks almost identical with the original circuit.
You also must add a 10 mfd 50 volt cap across the spreader for stability purposes. I feel that this cap should have always been there as it reduces circuit distortion just below clipping. I have had the amp running wide open on 8ohm dummy load for more than a day and it is completely stable. Sounds pretty good on music also when hooked up to my McLarrens. I have been tracking the two channels closely as the left channel is still all original.
Please See the images below.....
I hope that someone will find this helpful. I feel the effort to keep this great amp running was worth it.
All my best to all of you......🙂
You must add a second transistor wired as a thermal diode that gets attached to the heatsink in the stock location. It is wired in series with the original trans and effectively cuts the bias compensation by 50%. The existing trans is relocated to a bracket under the Main PCB which now only compensates for ambient temperature. The modified circuit now tracks almost identical with the original circuit.
You also must add a 10 mfd 50 volt cap across the spreader for stability purposes. I feel that this cap should have always been there as it reduces circuit distortion just below clipping. I have had the amp running wide open on 8ohm dummy load for more than a day and it is completely stable. Sounds pretty good on music also when hooked up to my McLarrens. I have been tracking the two channels closely as the left channel is still all original.
Please See the images below.....
I hope that someone will find this helpful. I feel the effort to keep this great amp running was worth it.
All my best to all of you......🙂
Last edited:
your circuits do not show, but it could just be me...
i also have this amp as a rebuild, i converted the traffo from 100 volts to 230 vots...
have you tried replacing the 6CG7 tubes with ECC88?
i also have this amp as a rebuild, i converted the traffo from 100 volts to 230 vots...
have you tried replacing the 6CG7 tubes with ECC88?
Hello Antonio, I haven't yet tried rolling any tubes. Have you ?
I have modified the post to include a note to right click and open in new window or tab and that seems to work. I guess the site won't allow you to open a direct link.
Also, I was wondering if you would allow me to edit the post now that my thirty minutes are up. I need to add one more image thumbnail at the bottom.
Cheers !
I have modified the post to include a note to right click and open in new window or tab and that seems to work. I guess the site won't allow you to open a direct link.
Also, I was wondering if you would allow me to edit the post now that my thirty minutes are up. I need to add one more image thumbnail at the bottom.
Cheers !
Hi wilsonkg, am in the process of updating a lv-105 that I totally recapped etc last year for a customer, now he would like me to change out working original outputs for the newer slightly more robust 'IRFP's' such as you mentioned above. In all the numerous articles on doing that, yours is the only one I've ever seen that fixed the higher Vgs of the IFRP's with adding the series diode wired c3423, and the 10uf cap. Was just wondering if you had a story to add to your idea of this mod, how did it work on yours, what bias amount did you end up setting the channels of your 105 to after adding the IFRPs and the c3423? Anyway, very curious about your mod and have seen these that just aren't very stable after adding new fets. Would love to try your mod. Thanks in advance, Steve
Hi Stevieg123,
I'm sorry but I don't have a magic story on how I cam up with the series diode idea. It actually came to me in my sleep but seemed logical (If the Vgs is wrong, correct it). The bias in this amp is set to factory spec which cam from the service manual. This amp is running beautifully and actually runs about 10 degrees C cooler than stock. I ran this unit on 4 and 8 ohm dummy loads at 1/2, 3/4 and full power for a whole afternoon before I decided that it was good enough to button up. The power output is about 2v higher with the higher current devices and measured out to about 90 watts continuous and the bias spreader curve is very close to original with these devices now. I originally upgraded one channel and then ran side by side comparisons. This is when i noticed the bias stability issue that others have noted. I added the diode junction and this worked great but I then noticed a strange distortion at just before full power. This didn't seem to be audible but looked like a knee on the scope. Adding the 10uf capacitor eliminated this and I could have probably gotten away with a lower value. Capacitors accross the Bias spreader is a fairly common approach.
I hope that if you give this a try that you will find it to be a worthy mod. I think the amp sounds better with the IFRP's and has a slight extension to the high's.
Cheers,
Kelly
I'm sorry but I don't have a magic story on how I cam up with the series diode idea. It actually came to me in my sleep but seemed logical (If the Vgs is wrong, correct it). The bias in this amp is set to factory spec which cam from the service manual. This amp is running beautifully and actually runs about 10 degrees C cooler than stock. I ran this unit on 4 and 8 ohm dummy loads at 1/2, 3/4 and full power for a whole afternoon before I decided that it was good enough to button up. The power output is about 2v higher with the higher current devices and measured out to about 90 watts continuous and the bias spreader curve is very close to original with these devices now. I originally upgraded one channel and then ran side by side comparisons. This is when i noticed the bias stability issue that others have noted. I added the diode junction and this worked great but I then noticed a strange distortion at just before full power. This didn't seem to be audible but looked like a knee on the scope. Adding the 10uf capacitor eliminated this and I could have probably gotten away with a lower value. Capacitors accross the Bias spreader is a fairly common approach.
I hope that if you give this a try that you will find it to be a worthy mod. I think the amp sounds better with the IFRP's and has a slight extension to the high's.
Cheers,
Kelly
Hello Kelly,
Am eagerly waiting my Fedx man to drop off my new IRFP's today, and I bought some 3423's off ebay the other day, should be right behind the fet's. So from my reading of all your info here, I can only decuce that what my plan of attack was will work out fine...do one channel first, entire mod including cap, leave factory bias setting alone on both ch's, and then do an A/B test between them. I may even have to pull out my little calculator and do a little math to determine current flow thru source resistors on old outputs and match the new ones as close to that as possible and then check bias tracking at a number of different load/level settings before I go 'bottoms up'. Will give you a review when I "get 'er dun". Thanks again,
Steve
Am eagerly waiting my Fedx man to drop off my new IRFP's today, and I bought some 3423's off ebay the other day, should be right behind the fet's. So from my reading of all your info here, I can only decuce that what my plan of attack was will work out fine...do one channel first, entire mod including cap, leave factory bias setting alone on both ch's, and then do an A/B test between them. I may even have to pull out my little calculator and do a little math to determine current flow thru source resistors on old outputs and match the new ones as close to that as possible and then check bias tracking at a number of different load/level settings before I go 'bottoms up'. Will give you a review when I "get 'er dun". Thanks again,
Steve
Hi Stevieg123,
I would like to point out a couple of safety items so that you don't suffer disaster like I did the first time around when I let the smoke out.
If you are using the exact parts that I did, make sure that R7301 is 3K. I also replaced all the resistors with Dale's at 1watt they were exactly the same size as the originals and shouldn't drift as they age. The other thing I did was use a 150 watt light bulb in series with the ac line when I first bring it up. This will give you safety as you will know as soon as the amp starts to pull excessive amounts of current and switch the bulb full on (Normally it will glow gently until you start to raise the volume or get the bias current too high). I would also start with the bias pot all the way to the counter clockwise position.
Do you have a service manual? If you don't, the correct way to measure the bias is across the two outer pins on the top of the three pin emitter resistors. The factory setting is 65mv.
Sorry that this is so long winded but, the final thing that I would like to mention is that before you start the mod, hook up your meter to the emitter resistor and get accustomed to how the bias comes on. It doesn't come on instantly like a traditional transistor amp and is effected by the dc offset of the driver tubes. You will see the bias voltage swing negative first and the climb to the positive and normal state. If your bias is set too high in the beginning to something like 200mv it will smoke the gates. This is also why the the 3K resistor is important. Also check the other resistors in the bias circuit to make sure someone hasn't subbed them wrong during a previous repair.
Ok, enough talk....
Good luck with the project and I hope that you enjoy the end result as much as I have. My best friend liked it so much that he bought one and wants it upgraded like mine.
Cheers,
Kelly
I would like to point out a couple of safety items so that you don't suffer disaster like I did the first time around when I let the smoke out.
If you are using the exact parts that I did, make sure that R7301 is 3K. I also replaced all the resistors with Dale's at 1watt they were exactly the same size as the originals and shouldn't drift as they age. The other thing I did was use a 150 watt light bulb in series with the ac line when I first bring it up. This will give you safety as you will know as soon as the amp starts to pull excessive amounts of current and switch the bulb full on (Normally it will glow gently until you start to raise the volume or get the bias current too high). I would also start with the bias pot all the way to the counter clockwise position.
Do you have a service manual? If you don't, the correct way to measure the bias is across the two outer pins on the top of the three pin emitter resistors. The factory setting is 65mv.
Sorry that this is so long winded but, the final thing that I would like to mention is that before you start the mod, hook up your meter to the emitter resistor and get accustomed to how the bias comes on. It doesn't come on instantly like a traditional transistor amp and is effected by the dc offset of the driver tubes. You will see the bias voltage swing negative first and the climb to the positive and normal state. If your bias is set too high in the beginning to something like 200mv it will smoke the gates. This is also why the the 3K resistor is important. Also check the other resistors in the bias circuit to make sure someone hasn't subbed them wrong during a previous repair.
Ok, enough talk....
Good luck with the project and I hope that you enjoy the end result as much as I have. My best friend liked it so much that he bought one and wants it upgraded like mine.
Cheers,
Kelly
Hi Kelly,
Sorry it's taken me so long to get back in here to update my progress, one thing after another, you know Murphys' Law...anyway, got all my semis in and got after it, changed out fets and added series 'diode' to bias circuit as well as 10 uf caps, made some brackets to hold original C3423's in the new position, under board like you did, pop riveted them to board and started to do ch to ch tests with 'curve tracer' device. I mainly use it to check semis but works nice locating a problem by comparing two points in one channel, then looking at the same two points in the other channel can quickly check for a bridged solder joint, or a missed soldering point. If the traces aren't exactly the same channel to channel, I know something isn't right. She looked all good so pluged her into VARIAC and switched both on, slowly raised a.c. input to Lux, current meter only twitching slightly as I increased juice, slowly, slowly, then...click, the wonderful sound of the speaker relay snapping on. Yeah!!! Did quick bias check to make sure that was good, and hooked up loads and sinewave to inputs and bingo, two great perfect looking 1kHz sinewaves on scope. Warmed up and slowly adjusted bias to get her to a place I felt it safe to unhook one leg of one of the 10ufs I added to see if I could see what you were seeing near clipping. Damn! That was really some ugly garbage/distortion just prior to clip, just like you'd seen I'd bet, before putting the caps in. Other channel still fine but one without cap...ugly ugly ugly stuff.
Put a few pix in to show work. WILL TRY TO FIGURE OUT HOW TO UPLOAD PIX SOON
Both channels biased up just fine, took a lot of small adjustments over a good 45 minutes to get it stable as heatsink takes quite awhile to reach operating temp. It (bias) does seem to track the heat better than with the original fet's now. This morning it came up with 65 and 66 mv across bias test points cold, so I could not be any happier with the way she's acting so far. I'm giving her a slow warm up now to take all the full power measurements in a bit so will get back with them and try to figure out uploading pix and talk soon. Thanks for all your advise and help,
Steve
Sorry it's taken me so long to get back in here to update my progress, one thing after another, you know Murphys' Law...anyway, got all my semis in and got after it, changed out fets and added series 'diode' to bias circuit as well as 10 uf caps, made some brackets to hold original C3423's in the new position, under board like you did, pop riveted them to board and started to do ch to ch tests with 'curve tracer' device. I mainly use it to check semis but works nice locating a problem by comparing two points in one channel, then looking at the same two points in the other channel can quickly check for a bridged solder joint, or a missed soldering point. If the traces aren't exactly the same channel to channel, I know something isn't right. She looked all good so pluged her into VARIAC and switched both on, slowly raised a.c. input to Lux, current meter only twitching slightly as I increased juice, slowly, slowly, then...click, the wonderful sound of the speaker relay snapping on. Yeah!!! Did quick bias check to make sure that was good, and hooked up loads and sinewave to inputs and bingo, two great perfect looking 1kHz sinewaves on scope. Warmed up and slowly adjusted bias to get her to a place I felt it safe to unhook one leg of one of the 10ufs I added to see if I could see what you were seeing near clipping. Damn! That was really some ugly garbage/distortion just prior to clip, just like you'd seen I'd bet, before putting the caps in. Other channel still fine but one without cap...ugly ugly ugly stuff.
Put a few pix in to show work. WILL TRY TO FIGURE OUT HOW TO UPLOAD PIX SOON
Both channels biased up just fine, took a lot of small adjustments over a good 45 minutes to get it stable as heatsink takes quite awhile to reach operating temp. It (bias) does seem to track the heat better than with the original fet's now. This morning it came up with 65 and 66 mv across bias test points cold, so I could not be any happier with the way she's acting so far. I'm giving her a slow warm up now to take all the full power measurements in a bit so will get back with them and try to figure out uploading pix and talk soon. Thanks for all your advise and help,
Steve
Hi Steve,
I'm happy to here that it has all gone well for you.
It's very gratifying for me when something that I have done can be helpful to others and I love to share in the experience.
It sounds like you have done a great job with your mods and I look forward to seeing those pictures.
I also love that fact that there are so many individuals out there that are keeping this great hobby alive.
Kelly
I'm happy to here that it has all gone well for you.
It's very gratifying for me when something that I have done can be helpful to others and I love to share in the experience.
It sounds like you have done a great job with your mods and I look forward to seeing those pictures.
I also love that fact that there are so many individuals out there that are keeping this great hobby alive.
Kelly
Hi Kelly,
Well, thank you. Without taking your time to put your ideas in print here, I'd still be pokin' around trying to come up with something like you did. I have no doubt you saved me an eternity by having your post here. Oh, by the way, I tried to figure it out but have no real clue how to upload pix to my post, any tips you can shoot me would also be very helpful, (I'm an old fart who can barely find the on/off switch on a computer, I spent way too much time in the 70's listening to Pink Floyd, waiting for disco to go away) it seems to me they want me to upload pix from a web address, not straight from my computer...is that at least right? Thanks again,
Steve
Well, thank you. Without taking your time to put your ideas in print here, I'd still be pokin' around trying to come up with something like you did. I have no doubt you saved me an eternity by having your post here. Oh, by the way, I tried to figure it out but have no real clue how to upload pix to my post, any tips you can shoot me would also be very helpful, (I'm an old fart who can barely find the on/off switch on a computer, I spent way too much time in the 70's listening to Pink Floyd, waiting for disco to go away) it seems to me they want me to upload pix from a web address, not straight from my computer...is that at least right? Thanks again,
Steve
Hi Steve, as the pictures go - here is what I know.
You might have noticed that in my original post that I have links to photos and then the same photos at the bottom of the post. Here is how I did it.
You must firstly post a reply using the full editor
To insert links to photos you will need an online storage account like dropbox. Create a folder in dropbox and upload your images into it. Then you right click on the image and select share public link. You then click the insert link button at the top of the editor and paste the URL from your clipboard into there and it will look like this.
https://www.dropbox.com/s/bt5yo51xpbvyds7/IMG_0354b.jpg
To add photos link directly you click the insert image button at the top of the editor and paste the URL from your clipboard into there and it will look like this.
Lastly if you want images directly in your post, you click the attachments button at the top of the editor and browse for the files there. This will upload your files directly to the message server and then you use the link provided and paste that into an insert image link.
Like this.
Hope this helps, and I hope that I haven't rubbed anyone the wrong way by explaining this here in this thread.
Cheers,
Kelly
You might have noticed that in my original post that I have links to photos and then the same photos at the bottom of the post. Here is how I did it.
You must firstly post a reply using the full editor
To insert links to photos you will need an online storage account like dropbox. Create a folder in dropbox and upload your images into it. Then you right click on the image and select share public link. You then click the insert link button at the top of the editor and paste the URL from your clipboard into there and it will look like this.
https://www.dropbox.com/s/bt5yo51xpbvyds7/IMG_0354b.jpg
To add photos link directly you click the insert image button at the top of the editor and paste the URL from your clipboard into there and it will look like this.
An externally hosted image should be here but it was not working when we last tested it.
Lastly if you want images directly in your post, you click the attachments button at the top of the editor and browse for the files there. This will upload your files directly to the message server and then you use the link provided and paste that into an insert image link.
Like this.
Hope this helps, and I hope that I haven't rubbed anyone the wrong way by explaining this here in this thread.
Cheers,
Kelly
Attachments
Hi Kelly,
Thanks for the info but here is what I read from reply:
Hi Steve,
9-2=92)(_$%^&23)W4653245^%)_+%&$^#5367326)(*$&8359-)($*^@$(_09133635(*)&%^*132%!
Cheers,
Kelly
I guess you all will have to take my word that I took pix of my work and it looks damn good!!! Ok, gonna go give her a look over, blow it out with some air, tie up any loose wire bundels, and go thru one more full heat cycle before I box her up to go back to California on Monday. Again, Kelly, thanks so much for all your help in this matter, and like you said, hopfully these exchanges between us wil help someone else doing a similar unit, these old Lux products are and will always be worth the time and trouble for their owners to keep 'em out of the recycle bin, always.
Steve
Thanks for the info but here is what I read from reply:
Hi Steve,
9-2=92)(_$%^&23)W4653245^%)_+%&$^#5367326)(*$&8359-)($*^@$(_09133635(*)&%^*132%!
Cheers,
Kelly
I guess you all will have to take my word that I took pix of my work and it looks damn good!!! Ok, gonna go give her a look over, blow it out with some air, tie up any loose wire bundels, and go thru one more full heat cycle before I box her up to go back to California on Monday. Again, Kelly, thanks so much for all your help in this matter, and like you said, hopfully these exchanges between us wil help someone else doing a similar unit, these old Lux products are and will always be worth the time and trouble for their owners to keep 'em out of the recycle bin, always.
Steve
i replaced the 6cg7 with a japanese 6rhh8 double triodes, good results....
my reasoning? higher transconductance and the fact that low rails were used...
sorry, i missed answering for a long time, but being the thread starter, you can always edit your post #1 as you see fit....let me know if you run into problems...
Still having trouble with images
Hi Tony, thank you for your response.
When I click on manage attachments I get the following message.
"The administrator has specified that you can only edit messages for 30 minutes after you have posted. This limit has expired, so you must contact the administrator to make alterations on your message."
Is there anything you can do?
Cheers,
Kelly
Hi Tony, thank you for your response.
When I click on manage attachments I get the following message.
"The administrator has specified that you can only edit messages for 30 minutes after you have posted. This limit has expired, so you must contact the administrator to make alterations on your message."
Is there anything you can do?
Cheers,
Kelly
- Status
- Not open for further replies.