Ok. If the input voltage is about 22V, and output is 12V I would equally divide the voltage difference between the LM reg and the gyrator transistor. That means about 5V across both of them, so the output of the LM317/337 must be set at 17V. That leaves 5V headroom for the gyrator, which should be plenty.
Ray
Ray
ok, cleaned up boards a liittle just to make sure no bad contacts i went over with soldering iron.
here is update on readings off components
+12v
d44h11
b 17.2
c 23.8
e 16.9
317
adj 23.4
out 23.9
in 25.3
2sk117
s 17.2
g 13.8
d 23.9
output 16.9v
-12v
d45h11
b -18.8
c -23.9
e -19.1
337
adj -23.5
in -25.4
out -23.9
2n5460
s -18.9
d -23.9
g -14.2
output -19.1
looks better, i think as far as consistency between both boards. But no dropout on 337 & 317 ?
here is update on readings off components
+12v
d44h11
b 17.2
c 23.8
e 16.9
317
adj 23.4
out 23.9
in 25.3
2sk117
s 17.2
g 13.8
d 23.9
output 16.9v
-12v
d45h11
b -18.8
c -23.9
e -19.1
337
adj -23.5
in -25.4
out -23.9
2n5460
s -18.9
d -23.9
g -14.2
output -19.1
looks better, i think as far as consistency between both boards. But no dropout on 337 & 317 ?
Ray
Thanks for the advice. According to the voltage divider formula, I can try various resistor combination values to reduce 5V to 2.5V. But which value is the correct ones and at what ratting should I choose.....1/4W 1/2W etc.....?
The rating can be small, power is not an issue here, so 1/4W or 1/8W is fine.
If you want to divide by two, you need two resistors with the same value. The divider loads the source IC, so the total value determines the current draw. You don't want to make that too big, but if you choose the resistors too high, the signal may sag because of the load of the target IC. So this will need a bit of experimenting
.Let's start with 0.5mA current, that shouldn't be a problem for most IC's. That means the total resistance has to be 10k (5V / 0.5mA). So each resistor is 5k or 4k7 for common E12 value.
Oh, I see...
You need to swap those around (R1 is between ADJ and OUT). Either you switched them, or there's a serious error in that excel sheet. These values should give you 18V out of the LM regs.Ray
Ray,
you mentioned r1 is between adj and out, but on my teddy board it is r2 that is in between adj and out. so I think my problem is in the excel sheet but I used the same excel sheet for the 5v teddy's I did and it came out perfectly fine.
Now, I think i've messed up the +teddy reg when I swapped the resistors around and back because I cant get nothing but 2v on the input and the 317 heats up real bad. From a 25v feed in gets me 2v on 317 in and the gyrator is not working?
If I cant fix this I think I'm gonna put the teddyreg on hold for a while at least until my summer classes are done as thing is taking too much time from my studies. what a bummer!
Thanks for all the help! maybe I'll try the raygulators seems to use less components so less can go wrong.
you mentioned r1 is between adj and out, but on my teddy board it is r2 that is in between adj and out. so I think my problem is in the excel sheet but I used the same excel sheet for the 5v teddy's I did and it came out perfectly fine.
Now, I think i've messed up the +teddy reg when I swapped the resistors around and back because I cant get nothing but 2v on the input and the 317 heats up real bad. From a 25v feed in gets me 2v on 317 in and the gyrator is not working?
If I cant fix this I think I'm gonna put the teddyreg on hold for a while at least until my summer classes are done as thing is taking too much time from my studies. what a bummer!
Thanks for all the help! maybe I'll try the raygulators seems to use less components so less can go wrong.
I'm using this schematic. If the print on the board is different, that sucks
. Just to be sure: you're not using a SuperTeddy board, right?Ok, first make sure the LM reg is outputting 18V. With R1 = 220 ohm and R2 = 3k this should be 18.3V to be exact. With the values for R3 and R4 that you have now, the output voltage should be around 13V then. Make sure R13 = R3 and R14 = R4.
If the output is still 18V, check for proper oriëntation of the transistors and measure the voltage at the junction of R3/R4. This should be 13V. If this is o.k., check the voltage at both sides of R6. This should be around 13V as well.
Ray
Last edited:
Looks like you have a Flea problem
As for TeddyReg, why not ask Teddy on PFM forums? Of course Ray is more than competent to find the bug
Yes Houston,
we've got a problem.
Can you tell me on your PCB wich XO's pin must have the upper soldering pad linked to the lower one? I've pulled the XO out (a pain!) and so for L1, defo a short between the XO's +5V and the ground on my pcb, can't see where and why!
Thanks,
Matthieu
I swear I'll be a good guy, please God leave me be in peace, I've paid the price many times now, have mercy!
we've got a problem.
Can you tell me on your PCB wich XO's pin must have the upper soldering pad linked to the lower one? I've pulled the XO out (a pain!) and so for L1, defo a short between the XO's +5V and the ground on my pcb, can't see where and why!
Thanks,
Matthieu
I swear I'll be a good guy, please God leave me be in peace, I've paid the price many times now, have mercy!
Last edited:
The rating can be small, power is not an issue here, so 1/4W or 1/8W is fine.
If you want to divide by two, you need two resistors with the same value. The divider loads the source IC, so the total value determines the current draw. You don't want to make that too big, but if you choose the resistors too high, the signal may sag because of the load of the target IC. So this will need a bit of experimenting
Let's start with 0.5mA current, that shouldn't be a problem for most IC's. That means the total resistance has to be 10k (5V / 0.5mA). So each resistor is 5k or 4k7 for common E12 value.
Ray
Hi Ray
This is the chip data sheet which indicates I in for the I2S will be feed
An externally hosted image should be here but it was not working when we last tested it.
It's +/-10mA
So from calculation for voltage divider I should choose 250R for R1 and R2 (5V/250R+250R)...........Is this right....???
No, I would not do that if I were you...
This rating is an absolute maximum, in other words: the chip will survive these values, anything more could destroy it. It's not a design parameter.
Look at the power supply voltage, it says 3.5V max. for Vd, but that doesn't mean you will supply it with 3.5V instead of the usual 3.3V?
Ray
This rating is an absolute maximum, in other words: the chip will survive these values, anything more could destroy it. It's not a design parameter.
Look at the power supply voltage, it says 3.5V max. for Vd, but that doesn't mean you will supply it with 3.5V instead of the usual 3.3V?
Ray
Flea fixed (the 100nF cap at the XO's PS lines...),
OsCon all around but DAC analog,
Super Reg à la Jung/Didden on both DAC analog and digital,
More depth and better image, all cleaner, neater, and that may sound strange to you guys, find the bass "lower" and/or less visible (Nordenstam "Welcome to Happiness" feats a very low wave, first I though it was gone but the end of the wave came like an earthquake and in a very low tone, lower than what I recall...), wich tends to invisible with that terrific crap of old british speakers. Does it make sens or it's pshyco-audio-idiot effect after mods? A doctor in the audience?
Also, the tuned filter in the DOS isn't supposed to have sighlty less highs? I'm missing nothing here.
Wonder if the fact main is on the low side (223V right now) can do something with bass, amp not being regulated I guess voltage is lower on power rails?
Next step, later, super reg at DOS.
OsCon all around but DAC analog,
Super Reg à la Jung/Didden on both DAC analog and digital,
More depth and better image, all cleaner, neater, and that may sound strange to you guys, find the bass "lower" and/or less visible (Nordenstam "Welcome to Happiness" feats a very low wave, first I though it was gone but the end of the wave came like an earthquake and in a very low tone, lower than what I recall...), wich tends to invisible with that terrific crap of old british speakers. Does it make sens or it's pshyco-audio-idiot effect after mods? A doctor in the audience?
Also, the tuned filter in the DOS isn't supposed to have sighlty less highs? I'm missing nothing here.
Wonder if the fact main is on the low side (223V right now) can do something with bass, amp not being regulated I guess voltage is lower on power rails?
Next step, later, super reg at DOS.
That's good news Matthieu! Now keep it singing!
If you chose the Butterworth filter for the DOS, there's no fall-off at 20kHz. The one with the constant group-delay is a Bessel filter and it has slight attenuation at 20kHz, but only 1 or 2 dB.
Lower mains voltage and thus lower V+ / V- should have little or no impact, as most amps have current sources inside to set the working point. Things like quiescent current should remain stable even if the mains varies a little. That's not the case with my tubeamp that is, but that's a different story... Or the amp must be a really crappy one
.Ray