RMI-FC100, a single stage audio power amplifier - Page 123 - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 7th November 2012, 12:49 PM   #1221
diyAudio Member
 
Join Date: Nov 2007
Location: Adelaide
Default Component woes!

I will admit that this project has gone way overtime in the build process, having purchased boards in the original group buy. In the past few days I have switched on one channel followed by the other a few days later.

In both cases no smoke, no drama and no instability, but biasing problems aplenty. I am using the NJL3281/1302 devices supplied by RS Components. The degree of matching was not good among the 10 of each device supplied, but that is not the whole storey. When first swiched on I had to bypass two thermal trak diodes to get the bias within range.

Originally I thought I may have been sold a faulty batch of devices, as it is well known that some early production thermal trak devices had the dies of the wrong diode installed in the device. However, on reading Rudi's experiences where he had to add a Schottky diode across one of the Ttraks as well as selecting the highest Vf Ttrak for bypassing; current thoughts are that I am up against a combination of "fudge factors"

Perhaps the chief suspect is the quality of the Toshiba 2SA/2SC devices supplied by Crickelwood in the UK. The PNP 2SA1360's look the part with the notched cutouts around the mounting holes. But the NPN 2SC3423's have a very "plain Jane" package which may well be fake. I would be interested in comments on the appearance of these devices from others. None of the devices met the suffix Y hFE spec' and several of them simply didn't work.

When powering up the second channel no amount of Ttrak bypassing or adjustment turned on the OP devices. Found that Q19 (2SC3423) had no base emitter junction so replaced it. Its Vbe does not seem a particularly good match to its complement which is not helping the general biasing problems.

Anyone tried any successful substitutes for the Toshiba devices?

For the front end power supply I am using a single loop C core (two coil bobbins) on which I wound two new bifilar secondaries using 0.6mm wire. The regulation is way better than necessary as the core is possibly rated for 60VA or more.

Keith
  Reply With Quote
Old 7th November 2012, 01:49 PM   #1222
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
My 1360/3423 are in different To126 packages. These came from Dalbani.
Once you have working transistors in those positions, they will not affect the quiescent biasing. The Vf of the diodes determines the biasing.
You can control the current through the biasing diodes by pacing a resistor across each diode. Try a 3k across one diode and remeasure the bias current. That 3k should pass about 0.2mA That current is now not passing through the diode and thus the diode Vf drops a little. You can cary on adding one 3k to each diode in turn and measure the effect of each modification.

You may find that you can get the bias voltage you need with this simple mod.

Try to keep the 500r near the lower end of it's adjustment range.
I don't think you want 2mA passing through 400r to 500r for a 0.8V to 1V of bias voltage that is not temperature compensated. But that is for later once you have a bias current in the correct range (37mV to 38mV across a pair of emitter pins). I have a 1.2mV range from highest to lowest (37.8mV to 39mV) with my selected pairs.
__________________
regards Andrew T.
  Reply With Quote
Old 7th November 2012, 02:19 PM   #1223
diyAudio Member
 
Join Date: Nov 2007
Location: Adelaide
Andrew, thanks for the mention of minimising pot R in the diode string. A bit analogous to "dead" capacitance in a condenser microphone or ES loudspeaker.

Maybe a question for Rudi_Ratlos. I thought it was agreed that when Iq of the driver transistors was raised by lowering R5 to 12R, C1 was unnecessary, yet Rudi's board seems to have resurrected it with that big pigtail component?

Keith
  Reply With Quote
Old 7th November 2012, 05:07 PM   #1224
DRZ1 is offline DRZ1  Canada
diyAudio Member
 
Join Date: Jan 2007
Location: Quebec City
I have the exact same problem with my second FC-100. I have to short two ThermalTrack diodes to be able to adjust the bias and the thermal compensation is not correct, got thermal runaway.

I will try with 5 diodes and bypass them with some resistors to see. The PNP thermalTrack have a Vf of 0.44V and the NPN one 0.62V ???? Is it normal ??? didn't check the spec sheet.

Last edited by DRZ1; 7th November 2012 at 05:09 PM.
  Reply With Quote
Old 7th November 2012, 05:55 PM   #1225
diyAudio Member
 
Rudi_Ratlos's Avatar
 
Join Date: Jun 2009
Location: Hamburg / Germany
Keith, the Vishay MKP1839 / 1F was introduced by Mihai starting in post #345.

I currently know 2 "reliable" sources for authentic Toshiba 2SA1360/2SC3423 transistors: BD Enterprises and Pacificsemi.

I built my prototype using ISC 2SA1360 and 2SC3423, and they did their job very well!
I am using authentic Toshibas currently.

I had the "ThermalTrak Vf" - problem, when I built my prototype and used NJL0302/NJL0281 transistors that I bought several years ago.
When I did the 2nd round of the FC-100 group-buy, I bought several hundred of NJLs from MOUSER (current batch) and while matching them,
I also measured the Vf of the TT-diodes.
None (!) of the diodes had a Vf > 560mV at 5mA /25 degrees.

Best regards - Rudi_Ratlos
Attached Images
File Type: jpg Toshibas.jpg (371.6 KB, 576 views)

Last edited by Rudi_Ratlos; 7th November 2012 at 05:59 PM.
  Reply With Quote
Old 8th November 2012, 07:44 AM   #1226
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally Posted by DRZ1 View Post
................................a Vf of 0.44V and the NPN one 0.62V ???? Is it normal ???..........
that is suspiciously high @ 2mA

Those NPN ones are the first I would add a resistor across. I would also take an NPN diode out of circuit so that you are using 3 low Vf and only 2 high Vf diodes in the string.

It may be that the NPN diodes will need a lower resistor value (to bypass more current) than the resistor fitted to the PNP.

If all the NPN diodes are Vf~600mV then a 680r resistor in parallel will bypass just under 1mA. The diode current is then just over 1mA. The Vf @ 1mA must be less than the Vf @ 2mA. Don't be surprised that the high Vf diodes need very low value resistors to get sensible bias voltage from the 5 diode string..

BTW.
I left all 5pin long on the top side of the PCB. I bent alternate in opposite directions to increase the gaps between adjacent pins. These long pins are now test points to attach clip on probes, or crocodile clips.
You can use these pins to clip across any diode to temporarily take it out of circuit. That then reduces bias voltage and bias current so that you can measure Vf when the sink is cold.

CAUTION !
Long PIN next to the fuse holder is very close. Be ultra careful you don't short Vdd to base PIN or Vee to diode PIN !!!!!!!!!!
__________________
regards Andrew T.

Last edited by AndrewT; 8th November 2012 at 07:55 AM.
  Reply With Quote
Old 8th November 2012, 01:35 PM   #1227
diyAudio Member
 
Join Date: Nov 2007
Location: Adelaide
[QUOTE=Rudi_Ratlos;3232413]Keith, the Vishay MKP1839 / 1F was introduced by Mihai starting in post #345.=QUOTE]

Indeed it was, but at post 573 it is suggested that when the driver transistor Iq is raised to around 100mA C1 is no longer needed. Users of Rudi's boards will be able to keep their options open as the post 345 R and C was a pain to accomodate on the original board.

We need to be carefull with terminology when discussing bias voltages to discriminate between device Vbe's and thermal trak diode Vf's. I would have thought the thermal trak diodes would be the same for NPN and PNP devices and that if they are at the same temperature the Vf should be similar. DRZ1's thermal trak figures are way more divergent than any I have measured. What temperature/running time are these being measured at? Maybe we are looking at some faulty production from Onsemi. Can't recall where I heard about the faulty production but had the impression it was not NPN or PNP specific.

Rudi, thanks for your comments on the Toshiba devices. Seem to have lost the knack of using quotes on this forum. Select the text, hit quote, then what?

Keith

Last edited by Keith Taylor; 8th November 2012 at 01:40 PM.
  Reply With Quote
Old 8th November 2012, 02:27 PM   #1228
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Select the post via the "quote" button/icon at the end of that post.

I can see you have accidentally removed [/ just before the last quote]
__________________
regards Andrew T.
  Reply With Quote
Old 9th November 2012, 01:19 PM   #1229
diyAudio Member
 
Join Date: Nov 2007
Location: Adelaide
Andrew, I tried your idea of putting a shunt resistor across a thermal trak diode instead of shorting it out and it works very well. I had two diodes shorted, so I removed the short on one and put what I thought was a 3k3 resistor in its place. I wound the pot to zero ohms and switched on. Without touching a thing I have near optimum bias on that channel.

It turns out that the resistor is 33K, so it just needed the slightest voltage reduction in the string to bring it in range. The OP device Iq's stabilise faster than with two diodes shorted. Good suggestion.

Keith

PS So far I have not soldered a single thermal trak device in place as it would reqire a very specially shaped bit to unsolder them. I cut off lengths of tinned copper wire, flatten one end by squeezing it with pliers and force it into the via. Before doing so you need to get the flat end at 90 degrees to your pliers or what ever tool you are using to insert them. Often they will come to rest against the bend in the device lead below the board. I think the reliability comes from the square corners of the device leads contacting the inside of the vias. It is rather time consuming I will admit.

Last edited by Keith Taylor; 9th November 2012 at 01:40 PM.
  Reply With Quote
Old 9th November 2012, 01:46 PM   #1230
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Roender found that adjusting the diode current he could alter the slope of the compensation with temperature, (at least that is what I remember of how I interpreted his reports).
He could use 4 or 5 or 6 diodes. Any of these could give the correct bias voltage simply by changing the diode string current.
He found that 2mA and 5diodes gave the best temperature compensation for his amplifiers with his transistors.

That does not mean that different diodes will always give the same result.
It appears that the 4 series of NJL have different diodes and that some from ONsemi may have had the wrong diode.

So some experimentation in diode current is probably advised, if tempco is not perfect.
Once the amp is up and running properly, at the desired Bias voltage/current, it is fairly easy to add different resistors, or no resistors, to see the effect on tempco.

What are the diode voltages? Is the 33k on the diode with the highest voltage? or another?
Depending on which you did, it may be prudent to add a resistor to another high Vf diode and dial in a little resistance with the 500r Vr.
__________________
regards Andrew T.

Last edited by AndrewT; 9th November 2012 at 01:51 PM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Single stage amplifier Skorpio Pass Labs 33 5th January 2011 02:03 AM
RMI-FC100 MKII, single ended folded cascode VAS roender Solid State 27 9th July 2010 10:56 PM
single stage common-drain amplifier Music_Sg Solid State 4 4th July 2006 05:45 PM


New To Site? Need Help?

All times are GMT. The time now is 01:44 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2