Hi Victor,
Just received mine today. Many thanks! Your packing job was exceptionally good.
Now I'll have to build a good power supply and see how pointless it is to measure the THD with the HP 339A.
-Chris
Just received mine today. Many thanks! Your packing job was exceptionally good.
Now I'll have to build a good power supply and see how pointless it is to measure the THD with the HP 339A.
-Chris
I probably over stressed the input stage with too much dcv when in dc coupled mode.
Shows nil 2H and too high 3H. Will have to replace protection fets and probably IPS pcb... they are matched pairs... so cant just change one bad ips.
THx-RNMarsh
Last edited:
Victor's oscillator works very well supplied from a LM317 setup as a current source.
There are two shunts which form the virtual ground on the oscillator PCB. This is what you are supplying.
You can get the details from Demian.
Last edited:
Hi David,
Thank you.
So just a CCS to feed the board and I can get the voltage compliance from Demian then? That sound right?
-Chris
Thank you.
So just a CCS to feed the board and I can get the voltage compliance from Demian then? That sound right?
-Chris
Use 4 9V cells and the LM317. There is a whole thread on it and a PCB design for it. The current setting for the LM317 is there somewhere as well.
What I discovered is that I can't start the oscillator with a 600 Ohm load with the current setting but once operating its all fine.
What I discovered is that I can't start the oscillator with a 600 Ohm load with the current setting but once operating its all fine.
Victor's oscillator requires 35V. Demian did up a design for the guys on the forum a few years ago. I thought it might save you some time to dup it. But it's not rocket for you.
A 38V raw supply should do. That's only 3V and ripple across the 317. You can play around with the current to find what works best.
I used three 12V gels which worked very well with the 317. So there you go 4-9V or three 12V gel.
Last edited:
CCS for Viktor
Since we are on the topic, there are a couple of things I always wanted to ask.
The Viktor has 2x TL431 on board configured to +/-15V, each fed by a 100R resistor.
The LM317 current source uses a 75R current setting resistor, so the CCS is ~17mA.
That means 3.4V drop across the two resistors, plus ~2.5V minimum headroom across the LM317, plus the 1.25V across the 75R current setting resistor.
We run out of voltage with 4x9V.
And then I am not sure how low noise a LM317 CCS is.
Since we only require 17mA or so, why not just use a FET or CRD as CCS ?
Patrick
Since we are on the topic, there are a couple of things I always wanted to ask.
The Viktor has 2x TL431 on board configured to +/-15V, each fed by a 100R resistor.
The LM317 current source uses a 75R current setting resistor, so the CCS is ~17mA.
That means 3.4V drop across the two resistors, plus ~2.5V minimum headroom across the LM317, plus the 1.25V across the 75R current setting resistor.
We run out of voltage with 4x9V.
And then I am not sure how low noise a LM317 CCS is.
Since we only require 17mA or so, why not just use a FET or CRD as CCS ?
Patrick
Last edited:
because we all have drawers full of 317's ?
and its good enough when combined with the rest of the board. certainly works fine on mine.
Alan
The Viktor will also work without the CCS when just connected to 4x 9V.
I wonder whether the 317 CCS still functions as intended when the dropout voltage is only 1.5V or less.
And how about a LT3092 CCS, with a minimum dropout of 1.5V ?
Patrick
I wonder whether the 317 CCS still functions as intended when the dropout voltage is only 1.5V or less.
And how about a LT3092 CCS, with a minimum dropout of 1.5V ?
Patrick
Strap a cap across the load and noise is gone.
470uF is what I used in my SVO. Zener shunt feed by 317. Works great.
Maybe the input voltage should be higher. 3 x12V gel is more like 39V and rechargeable.
I scored a bunch of 12V gel at work. Retired from the alarm system and a couple out of the emergency lighting.
470uF is what I used in my SVO. Zener shunt feed by 317. Works great.
Maybe the input voltage should be higher. 3 x12V gel is more like 39V and rechargeable.
I scored a bunch of 12V gel at work. Retired from the alarm system and a couple out of the emergency lighting.
Hi Demian,
Thanks, I'll go look for the thread.
-Chris
Thanks, I'll go look for the thread.
David, that is exactly how I envisioned running it. Thanks. I was also thinking of using gel cells. I guess that a charge controller should be used for the series string of unknown gel cells. Either that or charge each one separately (the same thing).
-Chris
Hi Demian,
I can't find anything under "1audio" as a thread for this. It is possible that the search only goes so far back. I tried looking for posts > a year ago and still didn't come up with anything.
Could the thread have been started by someone else by chance? I'll hunt given some information.
-Chris
I can't find anything under "1audio" as a thread for this. It is possible that the search only goes so far back. I tried looking for posts > a year ago and still didn't come up with anything.
Could the thread have been started by someone else by chance? I'll hunt given some information.
-Chris
Cheep way is to charge the gels from a current source with the batteries in parallel.
It's not the best but it works. For string charging you need to build a balance circuit to maintain quasi voltage across the gels. Like we do in out UPS.
I did not start the thread on a PCB for Victor's oscillator. Its starts here: http://www.diyaudio.com/forums/equi...-victors-low-thd-oscillators.html#post4164497 I need to open mine to see how i kludged it. Essentially i bypassed the two 100 Ohm resistors. The Lm317 + no resistors maximized battery life. important for 9V batteries which aren't cheap. I bought a pile from China via eBay. Years ago I visited the China factory making them (Only one then, probably still). They had every known brand on display since they made something like 80% of the world's 9V batteries. No good reason to buy premium batteries for this application.
I still get a significant improvement in noises using battery power. I have tried low noise linear and switching supplies but the power line to ground leakage is limiting performance.
I still get a significant improvement in noises using battery power. I have tried low noise linear and switching supplies but the power line to ground leakage is limiting performance.
Hi Demian, Essb,
Many thanks. This will help a lot.
Demain, have you considered using an isolation transformer? Just a thought. I have one here that I can try. I would really like an AC powered instrument since it will only live on my bench. Although I could go both ways with a switch so I can float it for situations where I do need the full performance. That should be a good balance between battery life and convenience.
It's interesting to note that the HP 3580A and 3581C both will run off of batteries. I should refurb one of them and check out the performance. Just curious about them.
-Chris
Many thanks. This will help a lot.
Demain, have you considered using an isolation transformer? Just a thought. I have one here that I can try. I would really like an AC powered instrument since it will only live on my bench. Although I could go both ways with a switch so I can float it for situations where I do need the full performance. That should be a good balance between battery life and convenience.
It's interesting to note that the HP 3580A and 3581C both will run off of batteries. I should refurb one of them and check out the performance. Just curious about them.
-Chris
I have even tried ultra isolation transformers and all it takes is a few pF of coupling to show up in the output. All below -140 dBV. I agree that a line powered solution would be preferable but at this level its almost impossible. Even looping the remote DC may cause too much leakage.