| Author |
Message |
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
 Posted: Sat Jan 26, 2008 7:29 am Post subject:
|
  |
|
| Etaoin wrote: | Ian, you mentioned the J109 above. Futurlec sell the J108 which appears to be the similar but with a larger Idcc (80mA for the J108) and lower Rds.
Both have a Ciss of 85pF though. |
Might not be too bad. According to the table from Fairchild I have, J108,109,110 are all from the same process. For some reason this table does not give the capacitance for the Jxxx devices, so I wasn't sure about that part. The 3 nA leakage will limit you at the low-frequency end, but it's worth a shot.
They also have J111, which might be worth a try.
I also remembered that some folks like the PN5432, which I just noticed is also from the same process as the Jxxx. And you could check for the PN4856, which is reasonably close to the PN4391. You might even find the PN4859, which is what Terry used originally. Haven't seen that around here, but who knows, maybe someplace over there carries it.
Ian |
|
|
Back to top
|
|
 |
andrewF
Joined: Dec 29, 2006
Posts: 1176
Location: australia
Audio files: 4
|
| Posted: Sat Jan 26, 2008 8:41 am Post subject:
|
|
|
Ian
What is the result of using a fet with a low Idss?
I have always used 2SK30 which has a whopping max Idss of 6.5mA (Vds = 10V).
Many of the Japanese synth-diy builders use them too (Takeda San introduced me to the 2SK30). Nobody seems unhappy with them.
I looked up the 2N4391 datasheet and saw a max Idss of 150mA.
Would my VCOs sound/perform differently or 'better' with a more suitable fet installed? |
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
| Posted: Sat Jan 26, 2008 10:20 am Post subject:
|
|
|
| andrewF wrote: | What is the result of using a fet with a low Idss?
I have always used 2SK30 which has a whopping max Idss of 6.5mA (Vds = 10V).
Many of the Japanese synth-diy builders use them too (Takeda San introduced me to the 2SK30). Nobody seems unhappy with them.
I looked up the 2N4391 datasheet and saw a max Idss of 150mA.
Would my VCOs sound/perform differently or 'better' with a more suitable fet installed? |
Well, the way I look at it, the Idss is (roughly) a measure of the maximum discharge current. It doesn't do you much good to have a small Rds resistance if the current is limited by another mechanism.
As far as the 2SK30, well now we are in MOSFET territory -- a whole nother topic. Since it is a depletion-mode device, you can indeed just pop it into Terry's design. But if you look at the data sheet you will see that again it is intended as a small-signal RF device. But I wouldn't argue with Takeda San, and if it works for you, then great!
I have a design for a super fast core that uses an enhancement mode MOSFET (VN0104 or BS170), which allows the LM311 comparator to be powered from ground, which improves the switching time. I developed a special pulse shaping circuit to discharge the cap in 200 ns with minimal ringing. I had this up on my website temporarily a couple of years ago. It's a screamer! I see now that I used a 2SC1583 pair for the converter, and according to my notes I got perfect tracking up to 40 kHz.
Here's a picture of the reset part of the waveform. The triangle is the switching pulse applied to the MOSFET and the other trace is the Saw .
You won't probably hear much difference with a fast reset, it just makes tracking more accurate.
Ian
| Description: |
|
| Filesize: |
18.36 KB |
| Viewed: |
10436 Time(s) |
|
|
|
|
Back to top
|
|
|
adhdboy
Joined: Feb 16, 2005
Posts: 57
Location: denver
|
Posted: Sat Jan 26, 2008 12:00 pm Post subject:
LM394 or SSM2210?
Subject description: I will definitely check out Ian's suggestions. |
|
|
I have the greatest respect for Ian. He is very analytical and makes excellent suggestions. My results (regarding LM394 versus SSM2210) were found informally by replacing my LM394 (can type) pairs with SSM2210 (DIP type). The SSM2210 seemed to just work after I popped it in replacing the LM394 and thus I suggested it's use instead of the LM394. I will definitely check out Ian's suggestions regarding the LM394 since it is more economical than the SSM2210. From reading Ian's posts and replies on the synth-diy list it is obvious that he has a lot of skill at in-depth electronic analysis (which is not my strongest suit) including the math behind it all. As I have often said "I'm a breadboard experimental type of guy" and I make suggestions based on observations rather than analysis. If Ian makes a suggestion it's definitely worth checking out.
_________________
There 10 kinds of people in this world that understand binary those that do and those that don't. |
|
|
Back to top
|
|
|
etaoin
Joined: Jun 30, 2005
Posts: 761
Location: Utrecht, NL
|
| Posted: Sat Jan 26, 2008 12:06 pm Post subject:
|
|
|
| frijitz wrote: | | They also have J111, which might be worth a try |
I think I might even have one of those in my parts bin...
_________________
http://www.casia.org/modular/ |
|
|
Back to top
|
|
|
dar303
Joined: Jul 15, 2007
Posts: 97
Location: Sweden
|
| Posted: Sat Jan 26, 2008 12:31 pm Post subject:
|
|
|
| I think I have a sh!tload of J109's if anybody in europe wants some. They are somewhere in my junk right now since I'm building a new workshop but I think I have like 1000 or so in unopened boxes! |
|
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005
Posts: 4127
Location: Mount Hope, KS USA
Audio files: 96
|
| Posted: Sat Jan 26, 2008 2:42 pm Post subject:
|
|
|
| Quote: | what about circuits that call for NPN/PNP pairs??
what is a good chip for those?
(i know there is one circuit that calls for one in the thomas henry vco book) |
 I can't find any reference in an of Thomas' books where he requires a matched NPN/PNP pair. Which circuit are you talking about?
_________________
My Site |
|
|
Back to top
|
|
|
Tim Servo
Joined: Jul 16, 2006
Posts: 924
Location: Silicon Valley
Audio files: 11
|
| Posted: Sat Jan 26, 2008 3:58 pm Post subject:
Question about transistor pairs - split off |
|
|
| Scott Stites wrote: | | Quote: | what about circuits that call for NPN/PNP pairs??
what is a good chip for those?
(i know there is one circuit that calls for one in the thomas henry vco book) |
I can't find any reference in an of Thomas' books where he requires a matched NPN/PNP pair. Which circuit are you talking about? |
That would be the 566 VCO. Known to some as the "Sheboygan." 
By the way, what about the 2SA798 PNP pair? This is recommended for Thomas' LM VCO and available from
www.electronicsurplus.com
www.electronix.com
www.talonix.com
and maybe others...
Tim (gonna go get a matched pair of mochas at Starbucks) Servo |
|
|
Back to top
|
|
|
Thomas Henry
Joined: Mar 25, 2007
Posts: 298
Location: Southern Minnesota
Audio files: 2
|
| Posted: Sat Jan 26, 2008 4:31 pm Post subject:
|
|
|
While it's true I have used Terry Mikulic's PNP/NPN converter in several circuits in the past, I've always indicated that the thing to do is to simply glue two discretes together. I'm not sure I've ever seen a monolithic mixed type, and certainly wouldn't expect it to be any good compared to the LM394.
Instead, use the Mikulic converter in circuits where super accuracy isn't required---percussion comes to mind---and just use two transistors.
Thomas Henry |
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
| Posted: Sat Jan 26, 2008 5:00 pm Post subject:
|
|
|
| Thomas Henry wrote: | While it's true I have used Terry Mikulic's PNP/NPN converter in several circuits in the past, I've always indicated that the thing to do is to simply glue two discretes together. I'm not sure I've ever seen a monolithic mixed type, and certainly wouldn't expect it to be any good compared to the LM394.
Instead, use the Mikulic converter in circuits where super accuracy isn't required---percussion comes to mind---and just use two transistors. |
IIRC Terry copied that from the old Moog designs. I've never understood what was meant by "matching" the transistors. Certainly their hFE's will be quite different. So you could match Vbe's at the same current level, I guess, and maybe this would give fairly good cancellation of the exponential prefactors. I'm too young to have worked with that method for VCO control, though.
You can buy complementary pairs in a single package from several manufacturers (just look in the Mouser catalog) but these are not monolithic, and I haven't ever seen anything about their matching. They would at least have the advantage of the two chips being in close proximity.
Another thing to look at "someday".
Ian |
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
| Posted: Sat Jan 26, 2008 5:19 pm Post subject:
Re: Question about transistor pairs - split off |
|
|
Yeah, I've used that quite a lot, especially where I need to drive several OTA's in parallel. In that case, performance is usually limited in the end by OTA non-idealities, so the 798 seems to work just fine as far as I have seen.
Ian |
|
|
Back to top
|
|
|
andrewF
Joined: Dec 29, 2006
Posts: 1176
Location: australia
Audio files: 4
|
| Posted: Sat Jan 26, 2008 6:31 pm Post subject:
|
|
|
| frijitz wrote: |
As far as the 2SK30, well now we are in MOSFET territory -- a whole nother topic. Since it is a depletion-mode device, you can indeed just pop it into Terry's design. But if you look at the data sheet you will see that again it is intended as a small-signal RF device. But I wouldn't argue with Takeda San, and if it works for you, then great!
I have a design for a super fast core that uses an enhancement mode MOSFET (VN0104 or BS170), which allows the LM311 comparator to be powered from ground, which improves the switching time. I developed a special pulse shaping circuit to discharge the cap in 200 ns with minimal ringing. I had this up on my website temporarily a couple of years ago. It's a screamer! I see now that I used a 2SC1583 pair for the converter, and according to my notes I got perfect tracking up to 40 kHz.
Ian |
 I won't argue with Takeda San too and will carry on using 2SK30 in my VCOs (at least until i finish my stash)
I have a copy of your Saw VCO (2005 - right?). 40 kHz! - that could be useful for a beat frequency style thru-0-VCO. |
|
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004
Posts: 24493
Location: The Netherlands, Enschede
Audio files: 298
G2 patch files: 320
|
| Posted: Sat Jan 26, 2008 6:45 pm Post subject:
|
|
|
| andrewF wrote: | | that could be useful for a beat frequency style thru-0-VCO. |
Would be an interesting subject for a new topic!
_________________
Jan
also .. could someone please turn down the thermostat a bit.
 |
|
|
Back to top
|
|
|
loss1234
Joined: Jul 24, 2007
Posts: 1536
Location: nyc
Audio files: 41
|
| Posted: Sun Jan 27, 2008 10:21 pm Post subject:
|
|
|
IAN-why did you take down the circuit you spoke of? it sounds like you put a lot of work into it. i'd love to see it. (as i have mentioned before, you must put out a book of your circuits! with theory. i would buy it!!!)
Scott/Thomas--sorry about the confusion. I didnt have the book in front of me at the time, i just remembered that when i was last reading it, there was a transistor pair that was not NPN. thanks for the clarification.
_________________
-------------------------------------------- check out various dan music at: http://www.myspace.com/lossnyc
http://www.myspace.com/snazelle
http://www.soundclick.com/lossnyc.htm http://www.indie911.com/dan-snazelle |
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
| Posted: Mon Feb 04, 2008 10:06 am Post subject:
|
|
|
| Etaoin wrote: | | Yes, but are any of the pcb designs out there actually using such careful designs? The ones I've tried (MFOS, CGS, Marc Bareille's ASM) seem to benefit from the SSM2210. |
I finally had a chance to look at your examples. We've already discussed the MFOS circuit. The other two are modifications of the EN 76 Saw VCO, originally designed by Terry Michaels.
This design does not even *have* a HF tracking adjustment! It was made to track by overcompensating the reset time correction to simultaneously correct for the converter's emitter-base resistance. This is a fixed correction. So it can only be expected to track properly if identical components to those of the original design are used!
If you want to use different components, then you need to readjust the tracking. You cannot make a meaningful comparison between different transistor pairs without doing this.
To me, it is very disappointing that every modification that has been made to Terry's original design has made it worse. It started with Bernie taking out the HF tracking adjustment and has gone downhill ever since. If you look at Terry's original drawing, you will see in big capital letters "DO NOT SUBSTITUTE ANYTHING!" There was a very good reason for that.
Ian |
|
|
Back to top
|
|
|
loss1234
Joined: Jul 24, 2007
Posts: 1536
Location: nyc
Audio files: 41
|
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
| Posted: Sun Apr 20, 2008 5:40 pm Post subject:
LM394 followup |
|
|
Back to the original question. I have just finished a VCO using an LM331 V/F converter IC and an LM394-based expo converter. This beastie tracks within 0.05% over 11 octaves up to 120 kHZ. Below 50 Hz it looses accuracy because of bias/leakage currents. These could probably be trimmed out to get a couple more octaves, but I won't be needing that.
I don't see the icon for flogging a dead horse, but I hope this will help convince everybody to stock up on the LM394.
Ian |
|
|
Back to top
|
|
|
loss1234
Joined: Jul 24, 2007
Posts: 1536
Location: nyc
Audio files: 41
|
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
| Posted: Sun Apr 20, 2008 6:47 pm Post subject:
|
|
|
| loss1234 wrote: | | anyway, any plans on releasing this vco? |
Stay tuned.
Ian |
|
|
Back to top
|
|
|
guitarfool
Joined: Feb 26, 2007
Posts: 161
Location: Maryland
Audio files: 8
|
| Posted: Wed Apr 23, 2008 6:09 pm Post subject:
Re: LM394 followup |
|
|
| frijitz wrote: | | I don't see the icon for flogging a dead horse, |
How about a sickly reindeer  
| frijitz wrote: | | but I hope this will help convince everybody to stock up on the LM394. |
Works for me. Futurlec has them (8 pin DIP) for $1.80 US in single quantities
http://www.futurlec.com/Linear/LM394Npr.shtml |
|
|
Back to top
|
|
|
CJ Miller
Joined: Jan 07, 2007
Posts: 368
Location: 127.0.0.1
|
| Posted: Sun Apr 27, 2008 12:59 am Post subject:
Re: LM394 followup |
|
|
| frijitz wrote: | | I have just finished a VCO using an LM331 V/F converter IC and an LM394-based expo converter. |
Does this work something like the ARP Chroma VCO? I was just reading Chamberlin's MAOM and wondering how this sort of thing works in real life! |
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007
Posts: 1734
Location: NM USA
Audio files: 54
|
| Posted: Sun Apr 27, 2008 8:44 am Post subject:
Re: LM394 followup |
|
|
| CJ Miller wrote: | | Does this work something like the ARP Chroma VCO? I was just reading Chamberlin's MAOM and wondering how this sort of thing works in real life! |
Yes, same principle. Except I'm running it at 10x the audio output frequency, followed by a string of digital dividers and a pair of 10-step waveform generators.
Ian |
|
|
Back to top
|
|
|
CJ Miller
Joined: Jan 07, 2007
Posts: 368
Location: 127.0.0.1
|
| Posted: Mon Apr 28, 2008 12:03 am Post subject:
Re: LM394 followup |
|
|
| frijitz wrote: | | CJ Miller wrote: | | Does this work something like the ARP Chroma VCO? I was just reading Chamberlin's MAOM and wondering how this sort of thing works in real life! |
Yes, same principle. Except I'm running it at 10x the audio output frequency, followed by a string of digital dividers and a pair of 10-step waveform generators.
Ian |
= =8O =XO
!!! |
|
|
Back to top
|
|
|
tonyc
Joined: Oct 11, 2007
Posts: 16
Location: london
|
| Posted: Mon Apr 28, 2008 2:46 am Post subject:
|
|
|
Hello
I know this thread is mostly about NPN pairs,
But has anyone used the 2SA1928 dual transistor.
I have tried it in some expo converters for Thomas Henry VCO1, and other
LM13700 type VCO, and seems ok.
Does it compare with Mat03 or others ?
thanks for any help
-tc |
|
|
Back to top
|
|
|
andrewF
Joined: Dec 29, 2006
Posts: 1176
Location: australia
Audio files: 4
|
|
|
Back to top
|
|
|
|
Forum index » DIY Hardware and Software
FAQ
Calendar
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
Chat Room
