Never mind.
I have some PFM teddyreg PCB's. Just need to get the components and build them up.
Teddy reg.
I have some PFM teddyreg PCB's. Just need to get the components and build them up.
Teddy reg.
Hi Mike,
You might actually be further ahead by using some RC filters to isolate each section from each other on these supply runs. This will reduce any noise and provide isolation. The actual voltage is not that critical. It's the amount of noise on the line and stability. You will have a little more variation in voltage, but less noise.
-Chris
These capacitors like to short. I wouldn't use them. Bypass with ceramic capacitors for the high frequency portion.
I never use capacitors this large on these regulators. Don't forget to install a reverse biased diode across the in and out of your regulators to prevent them from being destroyed when you turn the unit off.
You might actually be further ahead by using some RC filters to isolate each section from each other on these supply runs. This will reduce any noise and provide isolation. The actual voltage is not that critical. It's the amount of noise on the line and stability. You will have a little more variation in voltage, but less noise.
-Chris
Hi Mike,
Perfect.
Figure out the voltage drop you can live with and size the resistor for that. Your capacitor value is for high frequency noise, so lower value film caps work here. If you include an RC filter ahead of your regulator, you can nail the high frequency noise before it gets into the regulator.
I think you will get more mileage from this than different regulators as it addresses the weak area of most regulators. That is reduced high frequency performance. It's also more kind to your bank account.
-Chris
Perfect.
Figure out the voltage drop you can live with and size the resistor for that. Your capacitor value is for high frequency noise, so lower value film caps work here. If you include an RC filter ahead of your regulator, you can nail the high frequency noise before it gets into the regulator.
I think you will get more mileage from this than different regulators as it addresses the weak area of most regulators. That is reduced high frequency performance. It's also more kind to your bank account.
-Chris
Hi.
I checked the PSU of my amp (and CDP for that matter) and there are no RC networks on the PSU that I am aware of. Im suprised by this as they wouldnt be expensive for the manufacturer to implement and would by - all accounts - be beneficial to sound quality. So I'm up for adding this simple mod to my preamp PSU.
Regarding the voltage drop. I will read the LM4562 datasheet and see what the minimum voltage is for these to still work ok. However, am I right that opamps have an optimum supply voltage, and that to deviate from that could affect things such THD and PSSR? I dont want to worsen the peformance of the opamp in my attempts to limit noise on its supply. Any advice?
I'm also considering the possibility that the opamp may perform better on a lower voltage, such as 12V for example. Is this the case?
Also: Would the optimum soloution be to add RC filter neworks both before and after the regs? And what about the caps allready in place? Currently I have 1000uF before the reg and 220uF after, both good quality caps. If using an RC network do I need to consider changing these caps to accomodate the calculated values for the network? For example if the output filter required a 10uf cap then would I just have this on the regs output or would the network be better positioned after the 220uF cap allready there?
Slightly confused.
Mike.
I checked the PSU of my amp (and CDP for that matter) and there are no RC networks on the PSU that I am aware of. Im suprised by this as they wouldnt be expensive for the manufacturer to implement and would by - all accounts - be beneficial to sound quality. So I'm up for adding this simple mod to my preamp PSU.
Regarding the voltage drop. I will read the LM4562 datasheet and see what the minimum voltage is for these to still work ok. However, am I right that opamps have an optimum supply voltage, and that to deviate from that could affect things such THD and PSSR? I dont want to worsen the peformance of the opamp in my attempts to limit noise on its supply. Any advice?
I'm also considering the possibility that the opamp may perform better on a lower voltage, such as 12V for example. Is this the case?
Also: Would the optimum soloution be to add RC filter neworks both before and after the regs? And what about the caps allready in place? Currently I have 1000uF before the reg and 220uF after, both good quality caps. If using an RC network do I need to consider changing these caps to accomodate the calculated values for the network? For example if the output filter required a 10uf cap then would I just have this on the regs output or would the network be better positioned after the 220uF cap allready there?
Slightly confused.
Mike.
LM4562
Datasheet states operational voltage of 3.5 to 17v. Theres an awfull lot thats above my understanding so I cant see if using less that 15v will cause any problems or not. would dropping the voltage reduce the gain?
I need help on this one guys.
Mike.
PS: Is it worth addign an RC network to the poweramp PSU?
Datasheet states operational voltage of 3.5 to 17v. Theres an awfull lot thats above my understanding so I cant see if using less that 15v will cause any problems or not. would dropping the voltage reduce the gain?
I need help on this one guys.
Mike.
PS: Is it worth addign an RC network to the poweramp PSU?
Hi Mike,
How much headroom do you have in your 15 V supplies? In other words, how many volts before the regulators? I've already talked about this in post #233.
Now, it looks like you should have around 20 volts for your raw DC. How much ripple is there? You need to find out what the minimum voltage is in the "trough". I appears as there is probably not a great deal of room here as your minimum voltage needs to be 18 VDC to keep your regulators from dropping out. You could move to a low drop out type of regulator, but not yet. I'm thinking you could lose a volt as I built in some buffer by saying a 3 VDC input output differential minimum for these regulators. SO let's say you are dropping a volt. You need to figure out what the current draw is to find the value of resistance needed. Try a one ohm resistor in series to figure out the current draw.
-Chris
Edit:
How much headroom do you have in your 15 V supplies? In other words, how many volts before the regulators? I've already talked about this in post #233.
You need to add some voltage for low AC conditions as well. I am thinking you can reduce the input voltage a little bit which does two things for you. It allows the regulator to run cooler and it reduces the higher frequency noise on the raw, filtered DC. Your drop depends on how much current draw there is.
Now, it looks like you should have around 20 volts for your raw DC. How much ripple is there? You need to find out what the minimum voltage is in the "trough". I appears as there is probably not a great deal of room here as your minimum voltage needs to be 18 VDC to keep your regulators from dropping out. You could move to a low drop out type of regulator, but not yet. I'm thinking you could lose a volt as I built in some buffer by saying a 3 VDC input output differential minimum for these regulators. SO let's say you are dropping a volt. You need to figure out what the current draw is to find the value of resistance needed. Try a one ohm resistor in series to figure out the current draw.
-Chris
Edit:
No. The current draw occurs in sudden spikes and it's too high. You want a low resistance for any power stages.
I wonder which bit confused you.
you need to measure the voltage ahead of the voltage regulators and the mains voltage at that time.
Now, find out how low the voltage will go if the mains is at minimum supply voltage.
Will there be sufficient headroom to allow the regulator to perform without dropping out?
Many DMMs when set to ACvolts can measure the approximate ripple on a DC supply.
Set your meter to 200Vdc and measure your regulator input voltage.
If it's less than 20Vdc set your DMM to 20Vdc and measure the input voltage again.
Note this average DC voltage.
Set your DMM to 200Vac and measure the input voltage to the regulator. If it's less than 2Vac, then set your DMM to 20Vac and remeasure the voltage on the regulator input. If it's less than 1Vac then set your DMM to 2Vac and remeasure the ripple voltage on the DC supply.
The peak to peak ripple is approximately three times the DMM average reading.
Subtract half this ripple from the average DC voltage i.e. 18.5Vdc and 150mVac gives the minimum voltage in the troughs of 18.5 - [1.5 * 0.15] = 18.275Vdc.
This example will result in 3.275V across the regulator, but you need to find the voltage drop acrossthe regulator when mains supply is at minimum. This is where you subtract the correction you calculated right back at the beginning.
One final measurement. Try to measure the regulator input voltage when the system is drawing maximum current. Is this drop under maximum load significant? Then subtract that from the headroom across the regulator.
you need to measure the voltage ahead of the voltage regulators and the mains voltage at that time.
Now, find out how low the voltage will go if the mains is at minimum supply voltage.
Will there be sufficient headroom to allow the regulator to perform without dropping out?
Many DMMs when set to ACvolts can measure the approximate ripple on a DC supply.
Set your meter to 200Vdc and measure your regulator input voltage.
If it's less than 20Vdc set your DMM to 20Vdc and measure the input voltage again.
Note this average DC voltage.
Set your DMM to 200Vac and measure the input voltage to the regulator. If it's less than 2Vac, then set your DMM to 20Vac and remeasure the voltage on the regulator input. If it's less than 1Vac then set your DMM to 2Vac and remeasure the ripple voltage on the DC supply.
The peak to peak ripple is approximately three times the DMM average reading.
Subtract half this ripple from the average DC voltage i.e. 18.5Vdc and 150mVac gives the minimum voltage in the troughs of 18.5 - [1.5 * 0.15] = 18.275Vdc.
This example will result in 3.275V across the regulator, but you need to find the voltage drop acrossthe regulator when mains supply is at minimum. This is where you subtract the correction you calculated right back at the beginning.
One final measurement. Try to measure the regulator input voltage when the system is drawing maximum current. Is this drop under maximum load significant? Then subtract that from the headroom across the regulator.
Thanks.
Im not sure about all this to be honest. I have a very basic meter so I cant follow Andys tips.
Voltage before regs and bridge is 17v. So afeter rectification its around the 20v mark I think. Based on this I think Im safe going for a 1v drop. Im still unsure if I should add RC filters before and after the regs though?
Mike.
Im not sure about all this to be honest. I have a very basic meter so I cant follow Andys tips.
Voltage before regs and bridge is 17v. So afeter rectification its around the 20v mark I think. Based on this I think Im safe going for a 1v drop. Im still unsure if I should add RC filters before and after the regs though?
Mike.
Remember the CD63/7 mods Mike? I think you looked at that thread before.. Many people shorted the resistors that came after the 12v+- regulators because they harmed the sound.
Don't put any resistance after the regs.
BTW your Black Gate caps will already be lowering noise, and LM4562 are quite immune to noise so let's not take this too far
Simon
Don't put any resistance after the regs.
BTW your Black Gate caps will already be lowering noise, and LM4562 are quite immune to noise so let's not take this too far
Simon
Above is the Preamp PSU of the CA 340A.
Curently I have 220uF BG STD at C49/50. I have a pair of 1000uF Pana FC to replace C45/46. Diodes are shottkey DQ10's I think. Currently the positive rail has additional smoothing as the +5v microprocessor supply feeds off it. When I bought the 1000uF Pana FC caps I had the idea for seperating the +5v supply in mind. As Im not doing this now Im uncertain of replacing the 1000uF (+ rail) and 470uF (- rail) with the Panas is ok? Obviously replacing like for like value would be ok but If I want aditional smoothing am I better going for a 1500uF Pana FC on the + rail and 1000uF on the -?
Regarding the RC network. Should I add one before the regs only to stop any noise before it enters the reg (takins Simons advice into account)?
Mike
The 5Vdc supply has plenty of headroom.
try using a 10r and 1000uF in it's line. Leave the resistor legs long to clip on test leads.
If the peak voltage drop is too high then add a similar resistor in parallel to the 10r.
You might find that the peak voltage across the digital 10r is so small that you can afford to increase this to 22r or even 33r.
The same modification in both the 15V supplies should work and again trimming the resistance value to ensure adequate drop across the regulator during peak current demands.
The cap + resistor on the digital supply acts as a filter to spikes and glitches coming back up the line to feed through the audio regulators.
try using a 10r and 1000uF in it's line. Leave the resistor legs long to clip on test leads.
If the peak voltage drop is too high then add a similar resistor in parallel to the 10r.
You might find that the peak voltage across the digital 10r is so small that you can afford to increase this to 22r or even 33r.
The same modification in both the 15V supplies should work and again trimming the resistance value to ensure adequate drop across the regulator during peak current demands.
The cap + resistor on the digital supply acts as a filter to spikes and glitches coming back up the line to feed through the audio regulators.
Try adding the R-C before the regs but don't make any other changes at the same time. You need to hear if this is good or not, I can't tell you that as I've not tried. On paper it may reduce noise but listen to the mod and then decide.
If you've already sourced some Pana FC then use them before the regs, they'll be fine. They're low impedance types and I think good for smoothing because of this. They certainly don't sound as good as BGs next to things like op-amps but before regs they'll work nicely.
Don't worry about the value. Keep in mind Chris' advice about moderate upgrades as some inherent factors (like the switching chip, LM3886 circuit etc) will be the biggest barriers to ultimate sound quality.
Simon
PS - AndrewT's tip sounds good to put in too
If you've already sourced some Pana FC then use them before the regs, they'll be fine. They're low impedance types and I think good for smoothing because of this. They certainly don't sound as good as BGs next to things like op-amps but before regs they'll work nicely.
Don't worry about the value. Keep in mind Chris' advice about moderate upgrades as some inherent factors (like the switching chip, LM3886 circuit etc) will be the biggest barriers to ultimate sound quality.
Simon
PS - AndrewT's tip sounds good to put in too
Thanks guys.
I think I'll definately add an RC network on the +5v supply as it certainly wont hurt the sound qwuality and could improve it slightly. As for the +/- opamp supply, I'll try these Teddyregs from Pinkfish media. I allready have a set of PCB's and just need to buy the components to build them up. Voltage out is set manually by a resostor of pot, theyre dead easy to make on stripboard too. Alot of the guys on the forum have Naim equoppment and so far the reports have been pretty spectacular with regard to this reg. Its bascally a LM317 followed by a gyrator I think. Anyhow, Ive read nothing but extremely positive reports so far so Its worth a try. Using the teddyreg means the 220uF BG stds arent needed on the output so I can try them in the feedback on the power stages.
Mike.
I think I'll definately add an RC network on the +5v supply as it certainly wont hurt the sound qwuality and could improve it slightly. As for the +/- opamp supply, I'll try these Teddyregs from Pinkfish media. I allready have a set of PCB's and just need to buy the components to build them up. Voltage out is set manually by a resostor of pot, theyre dead easy to make on stripboard too. Alot of the guys on the forum have Naim equoppment and so far the reports have been pretty spectacular with regard to this reg. Its bascally a LM317 followed by a gyrator I think. Anyhow, Ive read nothing but extremely positive reports so far so Its worth a try. Using the teddyreg means the 220uF BG stds arent needed on the output so I can try them in the feedback on the power stages.
Mike.
Hi.
Getting back to mains filtering for a minuite:
I found a mod which is reported to be very good. Its basically using 2 x 0.47uF 600V Auricaps across L/N.
Details here:
Apparently this is a really effective mod so I might give it a go. If its good on the amp I'll try it on the CDP and maybe even add a pair to my mains filter inplace of the choke.
Mike.
Getting back to mains filtering for a minuite:
I found a mod which is reported to be very good. Its basically using 2 x 0.47uF 600V Auricaps across L/N.
Details here:
Apparently this is a really effective mod so I might give it a go. If its good on the amp I'll try it on the CDP and maybe even add a pair to my mains filter inplace of the choke.
Mike.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.