A case and power supply for Victor's low THD oscillators - diyAudio
Go Back   Home > Forums > Design & Build > Equipment & Tools

Equipment & Tools From test equipment to hand tools

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 27th December 2014, 04:13 AM   #1
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Default A case and power supply for Victor's low THD oscillators

I recently bought Victor's 1kHz and 10kHz oscillators after skimming through the long thread here that I haven't been following until now. I've spent a couple of hours with the thread so far but haven't seen anyone publish a case and power supply to use with his oscillators, so here goes since I need one myself. I'll post the Gerbers and FPE panel CAD file when the board is back and I've verified it works. If anyone has ideas or suggestions on ways to improve, or spots anything that I'm screwing up, please post!

The case is the extruded aluminum Hammond 1455N1601 (or 1455N1601BK for black or BU for blue) which is $24 at Mouser and Digikey right now with a bunch in stock. The case is designed for a 100mm wide by 160mm long PC board. I'm using a 2 layer here and so far I've been able to keep everything to the bottom layer. It would probably be possible to home-etch it, although I'm going to send it out for fab.

The case should be tall enough to fit both of Victor's boards together, as the front panel image below shows. His 10kHz board is on the bottom and his 1kHz board is upside down on the top. At 27mm tall each the board won't fit the 49mm tall case if stacked with both right side up, which is why the upside-down interleaving. The 1kHz is shifted to the left so the tall electolytics in the back of Victor's boards miss each other. The 160mm x 100mm power supply PC board is notched out around Victor's 10kHz oscillator board on the bottom. I tried fitting it all into a Box Enclosures B3-160 extruded case, also 100 x 160mm, but at 8mm less tall it doesn't appear the oscillator boards would clear in any configuration.

The power supply board sits in back and the left side and consists of 6 9V alkaline cells. The 5 in back sit on edge and wedge in between the back panel and a 9V PCB snap on the PC board. Pairs of cells have their top terminals bridged by 9V snap pairs with wire leads. The 6th cell is wedged between the case sidewall and the two PCB battery snaps. I have Schottky (for low Vdrop of course) diodes between all the cells to take into account someone accidentally touching the battery terminal with the battery backwards - lol I do it all the time. I know I'm going to have to find something with a higher PIV than the 40V of those 1N5819's. Any suggestions are welcome...

I'm using 6 cells since the 9V alkaline discharge curve here (scroll down) goes from 9V to around 6.5Vdc before dropping off a cliff. That gives 6.5Vdc x 6 = 39Vdc minimum, still above the 35Vdc target and a nice 4V across the Vreg. On the high side that would be 9Vdc x 6 = 54Vdc, which is well within the maximum 60V input spec for the LM317HV regulator or 125Vdc for the TL783 regulator. Turns out the LM317HV and TL783 have the same pinout and are the same mechanically, with the exception of the LM317HV tab being a tiny bit thicker.

The regulator is in the TO-220 package and bolted to the front panel for heat sinking. I've kept all the holes on the front panel so only one uber-expensive Front Panel Express panel is needed (their ordering program totals up $37.92 USD plus $5 shipping for the design). I've set the board up where it can either be wired as a 35Vdc voltage regulator or a 25 mA current source in the typical fashion. The notes on the schematic give the details. In the case of the current source the 100 R resistors on the power lines on Victor's board are jumpered, as suggested in the thread.

I'm using batteries to preserve the ultra low noise floor that Victor's oscillators apparently have from all the AP measurements in the thread. I originally was considering a voltage doubler AC supply with a 24 Vac transformer input. I think it would be difficult to keep the power line harmonics out of the oscillator, at those super low noise floor levels, if the AC supply were in the same box and without a metal shield around it. The current draw is low so the batteries should last a fairly long time. The four front panel mounting screws would have to come off and the whole assembly slid out to change the batteries.

There is one off-board panel mounted switch to send the power either to the 1 kHz oscillator or the 10 kHz. The power supply could easily power both at once. My thinking though is the two signals would get into each other and mix given the close proximity of the two oscillator boards here, especially facing each other like this with one upside down. Probably best to only have one oscillator board powered up at a time.

On the front panel I've specified a small 0.25 mm deep countersink around Victor's ground screws and one panel mounting screw to cut through the anodization for a good electrical contact. I probably should do the same around the pot shafts to use the pot nut and properly ground the shaft. The panel design is down with the free-download "front panel designer" on Front Panel Express (or Schaeffer AG in the EU, their parent company it seems) website. Using FPD the text or anything else on the panel can be easily modified.
Attached Images
File Type: jpg sig gen front panel 1.JPG (179.1 KB, 425 views)
File Type: jpg sig gen pwr supply layout.jpg (695.1 KB, 421 views)
File Type: png sig neg pwr supply schematic.png (26.1 KB, 399 views)

Last edited by agdr; 27th December 2014 at 02:25 PM.
  Reply With Quote
Old 27th December 2014, 11:05 PM   #2
diyAudio Member
 
SyncTronX's Avatar
 
Join Date: Oct 2008
Location: Tejas, where blues are king
Blog Entries: 3
Is that Ag Doctor?
Or Silver Doctor?

That looks rather nice. I think that it is kind of needed.

Paraphrasing here Demian (1Audio) can tell you better and me.
Demian has a model of Victors 1kHz oscillator using
the the LM317 and 4 batteries I think. He mentioned that
it doen't really need 35 volts to function correctly.

I'm not sure what he did...something about soldering the 317
directly to the 100 ohm SMDs or you put a resistor
across it. And if you don't switch it off it will still
last 30 - 40 hours.

I was going to try and do that with it and put is in a tea tin.

But, if you are having the boards made and the panel made
punched etc, I would be interested.

Cheers,
__________________
-- Sync
  Reply With Quote
Old 28th December 2014, 02:36 AM   #3
RNMarsh is offline RNMarsh  United States
diyAudio Member
 
RNMarsh's Avatar
 
Join Date: May 2012
Location: 2457 Cascade Trail; Cool, CA. 95614
I am interested also... need the punched panel very much.
But want the pcb also.

Would like a set that holds three osc, also.

-RMarsh
  Reply With Quote
Old 28th December 2014, 04:00 AM   #4
diyAudio Member
 
luvdunhill's Avatar
 
Join Date: Jul 2006
Interested.
  Reply With Quote
Old 28th December 2014, 04:20 AM   #5
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Quote:
Originally Posted by SyncTronX View Post
Is that Ag Doctor?
Or Silver Doctor?
Lol! "agdr" turned out to be my version of "asdf". When I started posting on the audio forums a few years ago I wanted a name completely different than what I use for a couple of other hobbies. Couldn't think of a thing though. So I just pecked out some random letters and came up with agdr. Then later on I learned the most commonly typed random letters are "asdf" so I wound up with two of the four at least.

I did see 1audio's post and that is exactly why the board can be wired up as a current source too. He had a very good idea there. Those TL431's on Victor's board work by shunting whatever additional current they need to (past what current the main circuit is consuming) to maintain 2.5Vdc at their reference pin, which in turn is set up as a voltage divider from the power rail. Things are usually designed so that at the main circuit's minimum load the TL431 is shunting maximum current, as long as that amount is within the 100mA max sink rating for the chip. Then when the main circuit is using maximum current the TL431 reduces its current flow to keep the reference at 2.5Vdc, and hence the rail voltage steady at 15Vdc by way of the ref pin voltage divider.

So long story short, the TL431 and the circuit can be fed from some fixed current source like this. The TL431's will sink whatever amount of that current the main circuit isn't using to maintain the 15Vdc rail voltages. The current source has a very high impedance looking back into it of course, better than the 100R TL431 cathode resistors. If the input voltage to Victor's board is 35Vdc and the two rails are 15Vdc, then the existing design current is (35Vdc - (2 * 15Vdc)) / (2 * 100R) = 25mA. Those two 100R TL431 cathode resistors on Victor's board (his R32 and R33) would be just shorted across under the board since they are not needed with the current source feed.

Or... leave the 100R resistors and wire the board up as the 35V voltage source and that works too.

Hey good to know there is some interest in the board/case! I'm going to make it for myself regardless, so might as well share. For the first run of boards I'll probably only do the minimum of 5 (Seeed Studio's fusion service) since there is inevitably some screw-up in the layout. But if there is interest I may get a list together after that and do a larger run, sold at-cost of course. From their webpage:

Fusion PCB, customize PCB prototype | Seeedstudio

20 boards 100 x 160 (have to use their 100 x 200 service) with HASL finish would be around $98 plus probably $12 or so shipping, or around $5.60 a board. I would probably go with ENIG finish though (gold) which adds a dollar or so per board.

Quote:
Originally Posted by RNMarsh View Post
I am interested also... need the punched panel very much.
But want the pcb also.

Would like a set that holds three osc, also.

-RMarsh
Good deal! I'm going to run a test panel first before posting the panel FPE CAD file to make sure it all works as planned.

Front Panel Express' price comes down a tiny bit in quantity too. Their free FPD computer program has the discount built in. At 5 boards it gives 5% and 10 boards is 10%. FPE also runs the occasional discount promo on their Facebook page. They had 20% off for a while last summer. They are pricey but do excellent work. You can also mail them the panel that comes with the case and they will use that for the stock. I've done that a few times with Box Enclosure cases. That gives nice anodized edges which otherwise wind up as bare aluminum when they mill it out of a bigger sheet of their own material.

Just curious, why three oscillator boards? I'll bet you are tuning one of them for something other than the stock 1kHz or 10kHz.


Below is another idea for the board. Same board but pads for a voltage doubler circuit are added under the batteries. 24Vac in gives around 67Vdc into the regulator, which would have to be the TL783 chip in that case. The LM317HV and TL783 have the same pinout and mechanical package so either will work in the board, but the TL783 has 125Vdc max input.

That way the same board could be built up with an AC supply if someone wanted, although I still think the power harmonics would probably get into the oscillator. I'm thinking might as well include that circuit since the fab costs the same either way and it wouldn't interfere with the batteries if not used. I'm also thinking of going to two panel mounted power switches, one for each board (1 kHz and 10 kHz), rather than the one switch to select one board or the other. That way if someone really did want to run them both at the same time they could.

That is a bidirectional transorb across the transformer input in the voltage doubler. Then capacitor pads for 7.5mm lead spacing at 16mm or 18mm diameter, which as low-Z/low-ESR can be a 2200uF 50V or 1000uF 100V these days at Mouser it appears. The resistors are just high-ohm safety bleeders across the caps.
Attached Images
File Type: jpg sig gen pwr supply layout.jpg (801.4 KB, 348 views)
File Type: png sig gen pwr supply schematic.png (29.6 KB, 326 views)

Last edited by agdr; 28th December 2014 at 04:47 AM.
  Reply With Quote
Old 29th December 2014, 03:14 AM   #6
diyAudio Member
 
SyncTronX's Avatar
 
Join Date: Oct 2008
Location: Tejas, where blues are king
Blog Entries: 3
agdr,

Thanks for the explanation.
Please count me in for the boards
and the front panel too.
Trying to drill out tea tins is kinda messy.

I like the functionality of the dual supply board
because if anyone else has my good luck I'd
just get it all set up dialed in then the batteries
would die. I'd go to replace them and be out
or not have enough. If needed it's always nice
to plug in.

Initially I was thinking it might be nice to have an
ON - OFF - ON, for either but if ever their was a
need to run both at once....

I'm wondering about front panel layout as one Oscillator
is upside down guessing this minimizes power supply wire
lengths. Limited middle space or end up three horizontal
boards 2 Victors plus yours.

Not sure how it would affect things if oscillator boards were
they were lined up vertically... | | Would simplify recognition
of the man-machine interface and components.
With small multiples or like multiples of the same.

It also might simplify scaling, that is: 1 victor's, 2 victor's, 3 victors'...more.

Maybe only slightly wider chassis.

Food for thought any way.
__________________
-- Sync
  Reply With Quote
Old 29th December 2014, 05:49 AM   #7
diyAudio Member
 
SyncTronX's Avatar
 
Join Date: Oct 2008
Location: Tejas, where blues are king
Blog Entries: 3
agdr,

That's not bad, someone could have come up with QWERTY.
What if Jay Kay Ell Sem? LOL

I'm not trying to knock heads or anything just trying my hand at
something that is easily recognizable and we don't need to think
to use it.

Using small multiples or like multiples everything just kind of
instinctively takes hold. Like analog clocks, gauges, meters,
read outs, etc.... You look once and get a feel for it.

We don't need to process .25 or .75.
Just by glancing at the clocks hands we know.

Of all the tasks that we do, writing, designing, coding,
manufacturing, layout, testing, training, presenting...

...the most difficult task

...is keeping it simple.

Richard, you thinking something along these lines?

I slammed this together:

Click the image to open in full size.
__________________
-- Sync

Last edited by SyncTronX; 29th December 2014 at 05:58 AM.
  Reply With Quote
Old 29th December 2014, 05:54 PM   #8
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
SyncTronX - good thoughts on the vertical mounting but the interior height of that Hammond 1455N1601 case is 49.6mm while Victor's board length is 59.75mm. So far that 1455N1601 case is the tallest extruded 100mm x 160mm with card slots that I've been able to find. If anyone knows of a taller case out there somewhere, please post!

The vertical height of Victor's boards would also prevent them both from being right-side up. They are 27mm tall, so make that 28mm with 1mm spacing between. 2 would be 56mm, more than that 49.6mm case height. By flipping one board over though it allows one board to encroach on the space of the other. The pot on one board sits a little bit above the RCA jacks on the other, and vice versa. Since the RCA jacks are short it all fits. Then moving one board to the left keeps those caps in the backs of his board from hitting each other.

Since there don't seem to be any taller 100mm x 160mm cases the best solution for 3 (or 4) boards is probably just to build a second case/power supply. I'll also post a front panel CAD with holes for just one of Victor's boards on the bottom, if anyone only wants to use just 1 of Victor's board (or a 3rd board in a second case).

I forgot to answer the question about the 6 batteries vs. 4. I'm designing it all for worst case, when the batteries run down after a long time and hit 6.5V each. 4 batteries would work as long as they are are fresh with voltage near the 9V or 8.5V end of things. From that discharge curve it doesn't look like "9V" batteries stay at 9V for long. They pretty quickly drop to 8.5V and then stay relatively steady in voltage drop over time from 8.5v down to 6.5V.

I would like to get some sort of "dead battery" indicator included to notify when the cells hit 6.5 x 6 = 39V. I've thought of a green LED in series with a 37V zener which would go off when the voltage was low, but I'm concerned about injecting any zener noise into the supply. If anyone has an idea on a low voltage indicator please post!

Last edited by agdr; 29th December 2014 at 06:00 PM.
  Reply With Quote
Old 29th December 2014, 10:17 PM   #9
diyAudio Member
 
SyncTronX's Avatar
 
Join Date: Oct 2008
Location: Tejas, where blues are king
Blog Entries: 3
Hi agdr,

I found some boxes at digikey that might work.
I selected 60.mm then got the other dimensions from
their drop down list.

http://www.digikey.com/product-search/en?pv329=771&FV=fff40021%2Cfff803a6&mnonly=0&newpr oducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid= 0&pageSize=25

I think they might have some other boxes too,
a variety of debths etc.

I see that you want it to lay flat front facing instead of
facing up. Some of the old Hard Drive boxes might work that
lay flat and have a tall enough face.

Also appreciate the PS explanation. I don't know enough about
it all. If you don't mind what are the benefits or advantages and
disadvantages of having a a voltage power supply vice current
power supply?

If you don't mind also, I'm struggling over at the HP339a Distortion
Analyzer thread, trying to figure out how to minimize 60 hz source
3rd harmonic modulation around the 1kHz notch filter. Last page is
where I'm currently. I've also used Victor's oscillator as an
input to the analyzer.

Cheers.
__________________
-- Sync
  Reply With Quote
Old 30th December 2014, 06:50 PM   #10
RNMarsh is offline RNMarsh  United States
diyAudio Member
 
RNMarsh's Avatar
 
Join Date: May 2012
Location: 2457 Cascade Trail; Cool, CA. 95614
I would like to have an extra for either low freq osc to really exercise the PS/amp and for a two tone IM test.

TH-Richard
  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
Low jitter crystal oscillators (NDK) Alexandre Swap Meet 1 20th February 2015 06:56 AM
Low jitter oscillators various frequencies jean-paul Swap Meet 2 28th April 2012 08:41 AM
preamp supply LOW power LOW noise LOW cost drdagor Power Supplies 1 14th September 2010 12:39 AM
DIY Audio Alchemy Power Station 2 Power supply PSU with case roscoeiii Swap Meet 2 29th May 2010 07:51 PM
using average oscillators for low distortion measurements unclejed613 Solid State 9 9th January 2007 05:36 PM


New To Site? Need Help?

All times are GMT. The time now is 09:23 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2015 DragonByte Technologies Ltd.
Copyright 1999-2015 diyAudio

Content Relevant URLs by vBSEO 3.3.2