RMI-FC100, a single stage audio power amplifier

Jump to Latest
A

AndrewT

R.I.P.
Joined 2004
post1016.
Scott Wurcer's comparison table seems to offer fair comparison.

The manufacturer's extracted data, shown in the post, uses different test conditions and cannot be used for comparison purposes.

The sk170 is designed and specified for audio use.
The f862 is designed and specified for RF use.
Yes, the f862 can be used for audio, but then it's audio performance should be compared to the sk170 jFETs.

What did Rick NL measure? I suspect he measured Idss. This is OK for building B1, but this is not matching.
http://www.diyaudio.com/forums/pass-labs/141162-2sk170-matching.html#post1785232

Rick refers to B1 on a chip
Even here Euvl commpares using manufacturer's data.
http://www.diyaudio.com/forums/pass-labs/140488-b1-turbo-chip.html#post1777801
 
Last edited:
A

AndrewT

R.I.P.
Joined 2004
AndriyOL said:
For En k170 and bf862 in its datasheets have a different measurement units. Don't know if they are crucially different.
Click to expand...
it's not the measurement units, they are the same.

It's the operating conditions that are different and these different conditions do result in different measurement data.

Look at the graph of transconductance vs Id for the f862 and sk170. They are almost superimposible. No where near the 2:1 difference quoted in the headline data shown.
 
Last edited:
A

AndriyOL

Member
Joined 2008
You're right. But I meant nV/Hz1\2 for sk170 and nV/VHz for bf862, if you understand what I'm about. I don't have these symbols on keyboard.
What you mean "superimposible"? Yes, Yfs under equal measurement conditions ought to be the same. In Scoutt Wurcer's table they are slightly different.
 
Last edited:
A

AndrewT

R.I.P.
Joined 2004
AndriyOL said:
You're right. But I meant nV/Hz1\2 for sk170 and nV/VHz for bf862, if you understand what I'm about.
Click to expand...
noise performance.
Yes the f862 is designed to be low noise at high frequencies.
The sk170 is designed to be low noise at audio frequencies.
The f862 does have good noise performance for an RF device right down to 1kHz and even pretty reasonable at 100Hz.

Superimpose = place one on top of the other. Scott recognised this and that's why he quoted the data he showed for comparison.

If you look at Id = 5mA and 10mA and 20mA you will see that the RF and AF jFETs perform almost identically for transconductance.
 
A

AndriyOL

Member
Joined 2008
AndrewT said:
noise performance.
Click to expand...
I'm about measurement unit. They are a little bit different for both devices regarding Hz in En.
AndrewT said:
The f862 does have good noise performance for an RF device right down to 1kHz and even pretty reasonable at 100Hz.
If you look at Id = 5mA and 10mA and 20mA you will see that the RF and AF jFETs perform almost identically for transconductance.
Click to expand...
I would say it has better Nf except at 10Hz.
It's good enough that bf at this parameter isn't worse than sk.
 
Last edited:
H

housing

Member
Joined 2006
AndriyOL,

If you want to use BF862 instead of 2SK170, try it yourself. Nobody has built a RMI-FC100 using BF in the front end so we can't help you out in this respect. Better still, build the 2SK170 version also and compare. Please don't forget to report back your findings here. :D

Thank you very much.

housing
 
Last edited:
A

AndriyOL

Member
Joined 2008
housing said:
AndriyOL,
If you want to use BF862 instead of 2SK170, try it yourself. Nobody has built a RMI-FC100 using BF in the front end so we can't help you out in this respect. Better still, build the 2SK170 version also and compare. Please don't forget to report back your findings here. :D
Thank you very much.
housing
Click to expand...

I'll report results without doubt, which have to be astonishing as I suspect, if I do this challenge. But I'm still collecting parts and don't have found output devices where to buy yet.
Maybe you know what value of grid stopper resistor shall I use (try to use) with bf or it will be ok with R27 2k?
Regards.
 
Last edited:
T

TedS

Member
Joined 2006
Sheldon said:
Ok, let's get started then. I recommend the front end supply as a good place to begin.

Sheldon
Click to expand...

Hi Sheldon,

It's taking me some time, but I'm trying to get spun back up on this. I've collected my boards and parts, and this is what I found:

Caps - I have everything except C1-4 for the PSU board. Someone suggested they were unnecessary, I guess.

Resistors - I have most, but not all.

Transistors - I have a matched set of J1 & J2 from AndrewT
I have 5 each of MJE15034/MJE15035, but need to match
I have ~15 each of NJL, but need to match

I do not have the BC550C, BC560C, 2SC or 2SA parts - I was waiting for someone who had said they would offer matched sets of these, but it appears they may no longer be interested.

Misc. - I don't have any diodes, pots, fuses, etc.

My plan is to simply start stuffing my boards with the parts I do have, and make a list of needed items as I go. I'll probably ask for input before ordering anything, and I'll need help with some of the matching. At least this way I'll build some momentum on the project. Any thoughts or suggestions?

Ted
 
A

AndrewT

R.I.P.
Joined 2004
start measuring your drivers and outputs.
That will keep you going till the 100's of BCs arrive. Then measure them until all the other components arrive.

Always start assembly with the lowest profile items. Zeners & signal diodes, then low wattage resistors, moving through the height range but leaving the outputs till very last.
Only solder the outputs after they have been screwed to the heatsink.
 
S

Sheldon

Member
Joined 2003
AndrewT said:
start measuring your drivers and outputs.
That will keep you going till the 100's of BCs arrive. Then measure them until all the other components arrive.

Always start assembly with the lowest profile items. Zeners & signal diodes, then low wattage resistors, moving through the height range but leaving the outputs till very last.
Only solder the outputs after they have been screwed to the heatsink.
Click to expand...

Good suggestions from Andrew. I don't think you need 100's of BC's, but they are cheap. If you order them in cut tape formate, you will find most of them to be quiet close. To match the outputs, I would put them in the board - no solder yet - and mount them to the heat sink. Then find some wire the right size to wedge the pins tight to the board, so that they will have good contact. Short one of the diodes with a clip lead and measure them in the circuit. Make sure you have them mounted well to the heat sink, as small differences in heat transfer will give you large differences in current. If you have a device that runs a little lower in current, compared to it's closest mates, use that one in the center position, as it will run slightly hotter.

Sheldon
 
D

DRZ1

Member
Joined 2007
Hi guys, last week a made a test on my second (not finished) FC100 amp.

I swapped the thermaltrack Motorola Transistors with Toschiba 2SC5200-2SA1943, and i liked it, warmer sound and a little more detailed.

I have to change the temp comp by something else to finish it. For the test i simply replaced the thermaltrack diode by BD139 BE jonctions, 4 was to much voltage an 3 it's under-compensated.

So i'm open to good solutions to make this mod to a already fantastic amp.
 
You must log in or register to reply here.
Top Bottom