Hi Andriyol,,
I know Canadian brands. For the price, a used 100 MHz is the way to go. Assuming a dual trace at least. The old HP analogue 'scopes are great. I rebuilt an HP 1722A from 1975 and it runs with the best today, really high quality stuff. Some of the olf Tek scopes are very good too. The Philips PM3070 has an amazing trace, but their is a small PCB (surface mount) that overheats. They also interfere with tuners. I bought one new and it needed service 4 times, but better than the equivalent Tektronics of the day. The HP / Agilent ones were already digital by then.
New DSOs can be good values. Keysight has a low line that is very good. Keysight is HP / Agilent these days. Rigol makes a reasonable scope.
Keep in mind that used scopes often need controls cleaned and power supplies repaired. One thing is true, spend as much as you possibly can on your 'scope. You're going to have it for decades. With digital, you need about 5X the frequency you will be looking at. Since FM is 120 MHz, I needed a 500 MHz scope. I got an Agilent 54642D and really love it. I find Keysight easier to deal with than Tektronix, and Keysight / Agilent / HP handed Tek it's butt in the scope wars. They are trying to catch up, but the gap is wide, competition is stiff.
-Chris
I know Canadian brands. For the price, a used 100 MHz is the way to go. Assuming a dual trace at least. The old HP analogue 'scopes are great. I rebuilt an HP 1722A from 1975 and it runs with the best today, really high quality stuff. Some of the olf Tek scopes are very good too. The Philips PM3070 has an amazing trace, but their is a small PCB (surface mount) that overheats. They also interfere with tuners. I bought one new and it needed service 4 times, but better than the equivalent Tektronics of the day. The HP / Agilent ones were already digital by then.
New DSOs can be good values. Keysight has a low line that is very good. Keysight is HP / Agilent these days. Rigol makes a reasonable scope.
Keep in mind that used scopes often need controls cleaned and power supplies repaired. One thing is true, spend as much as you possibly can on your 'scope. You're going to have it for decades. With digital, you need about 5X the frequency you will be looking at. Since FM is 120 MHz, I needed a 500 MHz scope. I got an Agilent 54642D and really love it. I find Keysight easier to deal with than Tektronix, and Keysight / Agilent / HP handed Tek it's butt in the scope wars. They are trying to catch up, but the gap is wide, competition is stiff.
-Chris
Hi quicknick,
I'm sorry, I guess you took that as a terse response from me. Not so, I'm in a reasonably decent mood today and nothing about your post changed that.
I was merely pointing out that you really didn't have any problems with regard to building at set of these amplifiers up. Not really, but it might involve more work than you are prepared for. Especially if you're uncomfortable with making your own circuit boards.
If you have the files used to make the boards (I think they were made available for download), you can have a board house make them for you. There is one that iis inexpensive from China you could try. I think they are "www.pcbway.com" if you want to check them out. Send files and some $$, get boards soon after. If you have a couple personal friends to go in with, they could be set up at the same time. I haven't used them yet, but I hear they are inexpensive and pretty quick for turnaround.
Sorry for giving you the impression I was in a bad mood.
-Chris
Edit: Get the parts before they become too difficult to find.
I'm sorry, I guess you took that as a terse response from me. Not so, I'm in a reasonably decent mood today and nothing about your post changed that.
I was merely pointing out that you really didn't have any problems with regard to building at set of these amplifiers up. Not really, but it might involve more work than you are prepared for. Especially if you're uncomfortable with making your own circuit boards.
If you have the files used to make the boards (I think they were made available for download), you can have a board house make them for you. There is one that iis inexpensive from China you could try. I think they are "www.pcbway.com" if you want to check them out. Send files and some $$, get boards soon after. If you have a couple personal friends to go in with, they could be set up at the same time. I haven't used them yet, but I hear they are inexpensive and pretty quick for turnaround.
Sorry for giving you the impression I was in a bad mood.
-Chris
Edit: Get the parts before they become too difficult to find.
Hi anatech,
Indeed it seemed terse to me, and since none of the other folks that asked for boards in this thread were answered to the line of "gtfo and build your own" I kinda wondered what part of my post warranted this. Anyway, I guess we can put this behind and I'm glad about the mood not being a problem
I understand part of your concerns, as I am uncomfortable to make my own PCBs. I'm not doing them for simple boards like soft-starts or speaker protection circuits, sure as hell I'm not going to build them for this amp, especially as they are double-sided (and the professionally-made PCBs do provide "that extra something", imo). However, the work involved doesn't scare me; if there'll be a future group buy I am prepared for this project to take a year or even more. I'm in no hurry as I have enough amps for my current needs (SymAsym, Elliott's P3a and P101, Gainclone - all built in the last decade, as a hobby, whenever the soldering iron fever touched me).
If there won't be another group buy, I guess that's it, I'll have to crown one of the above as my "magnum opus" , since a problem bigger than the boards is that of the hard-to-get parts. The 2SA/2SC transistors needed are not available anymore (not even at Mouser), and I'm not going to buy large amounts of NJL transistors in order to obtain the required matched groups.
So, thanks for the suggestions, now I think I've made my intentions clearer, if there will be another GB then I'm in, if not - shame on me for being idle all these years.
Indeed it seemed terse to me, and since none of the other folks that asked for boards in this thread were answered to the line of "gtfo and build your own" I kinda wondered what part of my post warranted this. Anyway, I guess we can put this behind and I'm glad about the mood not being a problem
I understand part of your concerns, as I am uncomfortable to make my own PCBs. I'm not doing them for simple boards like soft-starts or speaker protection circuits, sure as hell I'm not going to build them for this amp, especially as they are double-sided (and the professionally-made PCBs do provide "that extra something", imo). However, the work involved doesn't scare me; if there'll be a future group buy I am prepared for this project to take a year or even more. I'm in no hurry as I have enough amps for my current needs (SymAsym, Elliott's P3a and P101, Gainclone - all built in the last decade, as a hobby, whenever the soldering iron fever touched me).
If there won't be another group buy, I guess that's it, I'll have to crown one of the above as my "magnum opus"
, since a problem bigger than the boards is that of the hard-to-get parts. The 2SA/2SC transistors needed are not available anymore (not even at Mouser), and I'm not going to buy large amounts of NJL transistors in order to obtain the required matched groups.So, thanks for the suggestions, now I think I've made my intentions clearer, if there will be another GB then I'm in, if not - shame on me for being idle all these years.
Hi quicknick,
The matching doesn't have to be extremely close. At higher currents the emitter resistors force current sharing. At lower currents the matching helps reduce overall distortion. Matching the input pair is more important, and they aren't too expensive. The extra outputs could be used later if you have a failure. I don't think you need large amounts of transistors in order to create a matched set.
This is a really good amplifier, worth building. Same as your little Symasym was.
-Chris
The matching doesn't have to be extremely close. At higher currents the emitter resistors force current sharing. At lower currents the matching helps reduce overall distortion. Matching the input pair is more important, and they aren't too expensive. The extra outputs could be used later if you have a failure. I don't think you need large amounts of transistors in order to create a matched set.
This is a really good amplifier, worth building. Same as your little Symasym was.
-Chris
Hi Andriyol,
Guess why I keep an analogue scope on my bench as well as the $20,000 MSO (54642D new) on my bench? The current Keysight 'scopes are the best DSOs I have ever seen and they get better every year. The less expensive DSOs have some cool features, and they improve as well, but nothing gives the amount of detail or eye pattern clarity that an analogue scope does.
If you're going to go for a DSO, then I highly recommend that you buy a used Agilent or Keysight, take a small loan if you have to. Otherwise, that 5:1 ratio thing might bite you, and there is something called aliasing that can make signals appear that aren't really there. The other consideration is analogue bandwidth of the DSO. My Agilent does have bandwidth that supports the 500 MHz claim. Many don't.
Don't kid yourself, playing in DSO land gets expensive very quickly. Either that, or there are trade-offs that you might not become aware of for some time and you find out your instrument has been lying to you for years. Happened to me with a THD meter said to go up to 100KHz. Turns out it topped out at 30 KHz & I didn't discover this until I compared readings to an HP THD meter (they went to 1 MHz, and so captured all the harmonics). Embarrassing to say the least. An analogue scope used of good quality will pretty much show you the truth all the time. That's assuming your probes are of equivalent quality. I just threw out some and bought Keysight probes. More lies from equipment. Thankfully I didn't use them for anything important and expect very little from them.
If you buy something like a Rigol, at least you have something not that bad, but beware of the shortcomings of less expensive equipment. Note that I'm not saying you should buy something like mine (it was used, but still very pricey!). However, if you can swing something more expensive with a bank loan (not too crazy, a grand or two tops), you will be far better off.
-Chris
Guess why I keep an analogue scope on my bench as well as the $20,000 MSO (54642D new) on my bench? The current Keysight 'scopes are the best DSOs I have ever seen and they get better every year. The less expensive DSOs have some cool features, and they improve as well, but nothing gives the amount of detail or eye pattern clarity that an analogue scope does.
If you're going to go for a DSO, then I highly recommend that you buy a used Agilent or Keysight, take a small loan if you have to. Otherwise, that 5:1 ratio thing might bite you, and there is something called aliasing that can make signals appear that aren't really there. The other consideration is analogue bandwidth of the DSO. My Agilent does have bandwidth that supports the 500 MHz claim. Many don't.
Don't kid yourself, playing in DSO land gets expensive very quickly. Either that, or there are trade-offs that you might not become aware of for some time and you find out your instrument has been lying to you for years. Happened to me with a THD meter said to go up to 100KHz. Turns out it topped out at 30 KHz & I didn't discover this until I compared readings to an HP THD meter (they went to 1 MHz, and so captured all the harmonics). Embarrassing to say the least. An analogue scope used of good quality will pretty much show you the truth all the time. That's assuming your probes are of equivalent quality. I just threw out some and bought Keysight probes. More lies from equipment. Thankfully I didn't use them for anything important and expect very little from them.
If you buy something like a Rigol, at least you have something not that bad, but beware of the shortcomings of less expensive equipment. Note that I'm not saying you should buy something like mine (it was used, but still very pricey!). However, if you can swing something more expensive with a bank loan (not too crazy, a grand or two tops), you will be far better off.
-Chris
Hi Do,
You just need to match hFE, but the devices must remain at the same temperature. When it really matters, I have a jig that mounts all the devices together (up to 4) and I run the collector voltages (or drain voltages) up to the desired level. 10 V is normally enough, but I have run them up to 50 VDC or more. Then I run the base voltage up and with the aid of dropping resistors, each transistor is supplied with a similar base current. Then I allow them to warm up and settle in before measuring collector currents (built in 0R1 resistors). The transistors were matched for beta before this test.
No matter how you do this, it is in fact a physics experiment. Your patience and procedure will determine the outcome. Just use common sense and good judgement and you will be fine.
To measure beta on the devices, first lay them out so you can move your test leads from part to part without touching them. Then after they have settled in, measure the beta without touching them. Don't breathe on them either! A piece of metal across the top will stabilise them and help to equalise temperature. Do something like eat a meal after you set them up, or set them up and measure first thing in the morning. Then once you have selected the likely candidates, attach them to your jig and have them run at a current that makes them warm compared to your body and air temperature. It will be very uncomfortable in Death Valley! Once they have stabilised you can measure the collector (drain) voltage drops, and if you want to be very accurate, the base resistor voltage drops as well. From there you can figure out the static beta. If it is for an output stage, you can run them at the circuit operating voltages and bias currents. However, to be effective the heat sink should be small so it heats up more quickly with less cooling around the edges. This may preclude you testing the transistors at the expected operating current and voltage. All you really need to do is get them warm and stable, then your measurements will be valid.
-Chris
You just need to match hFE, but the devices must remain at the same temperature. When it really matters, I have a jig that mounts all the devices together (up to 4) and I run the collector voltages (or drain voltages) up to the desired level. 10 V is normally enough, but I have run them up to 50 VDC or more. Then I run the base voltage up and with the aid of dropping resistors, each transistor is supplied with a similar base current. Then I allow them to warm up and settle in before measuring collector currents (built in 0R1 resistors). The transistors were matched for beta before this test.
No matter how you do this, it is in fact a physics experiment. Your patience and procedure will determine the outcome. Just use common sense and good judgement and you will be fine.
To measure beta on the devices, first lay them out so you can move your test leads from part to part without touching them. Then after they have settled in, measure the beta without touching them. Don't breathe on them either! A piece of metal across the top will stabilise them and help to equalise temperature. Do something like eat a meal after you set them up, or set them up and measure first thing in the morning. Then once you have selected the likely candidates, attach them to your jig and have them run at a current that makes them warm compared to your body and air temperature. It will be very uncomfortable in Death Valley! Once they have stabilised you can measure the collector (drain) voltage drops, and if you want to be very accurate, the base resistor voltage drops as well. From there you can figure out the static beta. If it is for an output stage, you can run them at the circuit operating voltages and bias currents. However, to be effective the heat sink should be small so it heats up more quickly with less cooling around the edges. This may preclude you testing the transistors at the expected operating current and voltage. All you really need to do is get them warm and stable, then your measurements will be valid.
-Chris
Hi Chris,
Seems that I was really lucky with my first NJLs samples from ONsemi. All of them are very well matched, not only between the same polarity but also cross polarity. My presumption is that in the early production stages they had used MJWs dies for building NJL's zero series. Sadly they discontinued this process, but as you already mentioned cross matching is a nice to have but not as important as matching between the same polarity.
Happy building!
Mihai
p.s After 10 years listening to RMI-FC100 I still believe is the best I have ever build and there's nothing to further develop and improve to it.
Last edited:
Hi Mihai,
I found the same. I also really miss the MJW series I tested. Who knew they would be a flash in the pan?
Probably what happened was they didn't sell because people would buy the higher rated parts and completely miss what was great about the MJW parts. I really, really miss them.
Of course, how good the amp sounds does depend on how much effort is put into building it in the first place. It sure is a winner.
-Chris
I found the same. I also really miss the MJW series I tested. Who knew they would be a flash in the pan?
Probably what happened was they didn't sell because people would buy the higher rated parts and completely miss what was great about the MJW parts. I really, really miss them.
Of course, how good the amp sounds does depend on how much effort is put into building it in the first place. It sure is a winner.
-Chris
Guys, I need some help please...
As per my previous posts, one of my FC-100 module died some time ago. It kept burning PSU fuses so I checked it but couldn't find the issue. After chacking some more, I realized that most output devices had been destroyed. So I went ahead and ordered a bunch of N and P and matched them, soldered them to the PCB and changed the 15034-35 drivers.
Checked all voltages from shunt and PSU before, checked fuses and powered on. Amp powered properly and I started the calibration process, that is bias. I turned it all the way to 38mV (reading across output resistors) and then after a minute or two, D2 went very dim and bias dropped to 0mV... Heatsink got a little colder and then D2 came back fully lit and bias went back to normal. On next power up, D2 kept very dim but I also realized that it was also D7 that was dim and bias stays at 0mV.
I'm guessing something is wrong with the input stage, probably got damaged when the output stage got destroyed. I can check voltage reading but I don't know what to look for really... Could the pre-drivers be damaged as well? If it is the case, Q19, Q20 are obsolete and because of this, finding a match set is not going to be an easy task... I don't mind paying but I want to be sure it is the problem.
Thanks
Do
As per my previous posts, one of my FC-100 module died some time ago. It kept burning PSU fuses so I checked it but couldn't find the issue. After chacking some more, I realized that most output devices had been destroyed. So I went ahead and ordered a bunch of N and P and matched them, soldered them to the PCB and changed the 15034-35 drivers.
Checked all voltages from shunt and PSU before, checked fuses and powered on. Amp powered properly and I started the calibration process, that is bias. I turned it all the way to 38mV (reading across output resistors) and then after a minute or two, D2 went very dim and bias dropped to 0mV... Heatsink got a little colder and then D2 came back fully lit and bias went back to normal. On next power up, D2 kept very dim but I also realized that it was also D7 that was dim and bias stays at 0mV.
I'm guessing something is wrong with the input stage, probably got damaged when the output stage got destroyed. I can check voltage reading but I don't know what to look for really... Could the pre-drivers be damaged as well? If it is the case, Q19, Q20 are obsolete and because of this, finding a match set is not going to be an easy task... I don't mind paying but I want to be sure it is the problem.
Thanks
Do
Last edited:
Hi Do,
Too bad you weren't closer. I'd stick it on my bench with you here.
Input transistors get reverse biased (one does) whenever the output goes DC. If they were matched before, they may not be now. Also, may have damaged one of them.
Check solder connections on the tail current source. If that drops out, bias dies as well in most amplifier designs.
-Chris
Too bad you weren't closer. I'd stick it on my bench with you here.
Input transistors get reverse biased (one does) whenever the output goes DC. If they were matched before, they may not be now. Also, may have damaged one of them.
Check solder connections on the tail current source. If that drops out, bias dies as well in most amplifier designs.
-Chris
Hi Chris,
I wish I could be there in person, we could also have a good talk about many things I'm sure.
If you want, I could ship the module to you and you could let me know how much to fix it. Just a thought.
Feel very comfortable to say no
Thanks
Do