Hi,
the recommendation for 18Vac transformer input rather than 15Vac for a 15Vdc output voltage seems a bit high.
The minimum and maximum voltages coming from an 18Vac transformer are 24Vdc to 31Vdc.
After allowing for a Vdrop through 1r, regulator drop and 2V of ripple, these input voltage requirements still seem a bit high.
I realise these were based on real circuit measurements/simulations but would anyone care to comment?
BTW 15Vac gives 20Vdc to 26Vdc, in all cases less an allowance for diode drops of about 0.6V (2 shottky on low load) to 1.4V (2 silicon on full load). A single bridge would halve the diode losses.
Has anyone calculated/simulated the minimum regulator drop?
the recommendation for 18Vac transformer input rather than 15Vac for a 15Vdc output voltage seems a bit high.
The minimum and maximum voltages coming from an 18Vac transformer are 24Vdc to 31Vdc.
After allowing for a Vdrop through 1r, regulator drop and 2V of ripple, these input voltage requirements still seem a bit high.
I realise these were based on real circuit measurements/simulations but would anyone care to comment?
BTW 15Vac gives 20Vdc to 26Vdc, in all cases less an allowance for diode drops of about 0.6V (2 shottky on low load) to 1.4V (2 silicon on full load). A single bridge would halve the diode losses.
Has anyone calculated/simulated the minimum regulator drop?
Jens tested the regulator and saw that with 15VAC input the bottoms of the ripple showed up in the output at some loads - I think it was 1 A. This problem went away when he increased the input voltage with his variac. I chose 18V as the next highest voltage standard transformer.
I ran a series of simulations, and if the input voltage is 15V at .5A the bottom of the ripple voltage is barely above 17V. The regulator pass stage is a darlington. Therefore it needs 1.2V (2 x Vbe) plus a bit more for the transistors' C-E drop. In choosing capacitors, I looked for the voltage to remain above 18V going into the regulator.
As noted earlier, if the load is low, and the transformer is on spec (16.5V no load, 0R75 internal resistance) and 2700 uf or better is used in the filters then 15 V is sufficient. However, to allow for heavier loads and transformer production tolerance, I chose to go with 18V.
I had a similar issue with some PE buyout transformers. They are rated at 14V, measure 17.8VAC no load and are as big as AVEL's 100VA models. Using a 1n4007 bridge and 3900 uf I still had little dropouts on the output of my 7815 regulators powering 14 NE5532s. Going to 12V regulators cured the issue.
I ran a series of simulations, and if the input voltage is 15V at .5A the bottom of the ripple voltage is barely above 17V. The regulator pass stage is a darlington. Therefore it needs 1.2V (2 x Vbe) plus a bit more for the transistors' C-E drop. In choosing capacitors, I looked for the voltage to remain above 18V going into the regulator.
As noted earlier, if the load is low, and the transformer is on spec (16.5V no load, 0R75 internal resistance) and 2700 uf or better is used in the filters then 15 V is sufficient. However, to allow for heavier loads and transformer production tolerance, I chose to go with 18V.
I had a similar issue with some PE buyout transformers. They are rated at 14V, measure 17.8VAC no load and are as big as AVEL's 100VA models. Using a 1n4007 bridge and 3900 uf I still had little dropouts on the output of my 7815 regulators powering 14 NE5532s. Going to 12V regulators cured the issue.
I will look into the drop out voltage of the regulator.
What is the story on caps? what size is going into the kits?
I will rebuild the test setup to reflect the changes in the input filter.
If anyone has 50VA transformers, try testing the regulation and post some numbers. I don't know whats out there, but a safe bet is the 18VAC transformer, if someone has a great 15VAC transformer it might be ok, but from my point of view it is better to go for the 18VAC as this will work.
\Jens
What is the story on caps? what size is going into the kits?
I will rebuild the test setup to reflect the changes in the input filter.
If anyone has 50VA transformers, try testing the regulation and post some numbers. I don't know whats out there, but a safe bet is the 18VAC transformer, if someone has a great 15VAC transformer it might be ok, but from my point of view it is better to go for the 18VAC as this will work.
\Jens
A way to save space is to get rid of the series resistor in the input filter.
Doing this will free som space that I think could be put to better use - what do you think?
BTW testing with 2x1000µF in the input filter, just enabled operation from a 16VAC transformer @ 0.75A out.
If the series resistor is dropped I think (not tested yet) a 15VAC transformer could work ok.
\Jens
Doing this will free som space that I think could be put to better use - what do you think?
BTW testing with 2x1000µF in the input filter, just enabled operation from a 16VAC transformer @ 0.75A out.
If the series resistor is dropped I think (not tested yet) a 15VAC transformer could work ok.
\Jens
We weren't planning to have it scored, but there is no electrical connection between the two regulators. A hacksaw would make quick work of separating the two halves.
5V operation would require a different voltage reference (Yes, I know, the TL431 would work if the board was configured for it) A green LED would work, with the appropriate resistors in the voltage divider.
5V operation would require a different voltage reference (Yes, I know, the TL431 would work if the board was configured for it) A green LED would work, with the appropriate resistors in the voltage divider.
Latest Data here:
http://www.delta-audio.com/temp_jens/PSU_DIY_ver_1.7.zip
Have fun with it.
There is now room for 10.5mm output caps, and I removed the 1 ohm series resistor in the filter.
\Jens
http://www.delta-audio.com/temp_jens/PSU_DIY_ver_1.7.zip
Have fun with it.
There is now room for 10.5mm output caps, and I removed the 1 ohm series resistor in the filter.
\Jens
Thank you Jens, it looks great!
The BOM is an Eagle file, I pasted it into 'text' format for anyone who doesn't have Eagle.
Qty Value Parts
4 X1, X2, X3, X4
4 1µF C6, C12, C14, C25
12 1N4004 D1, D2, D3, D4, D5, D6, D11, D12, D13, D14, D15, D16
10 1k R1, R2, R4, R5, R10, R12, R19, R20, R21, R22
4 3k3 R6, R11, R13, R17
2 10V D8, D10
4 10k R3, R7, R16, R18
2 10k P1, P2
2 22k R8, R15
2 100R R9, R14
12 100nF C1, C2, C3, C8, C9, C13, C15, C16, C17, C21, C22, C26
2 100pF C7, C20
4 470uF 25V C10, C11, C18, C19
4 1200µF 35V C4, C5, C23, C24
3 BC546C T3, T4, T7
3 BC556C T2, T5, T6
2 GREEN 5mm LED1, LED2
2 LM4040D10 D7, D9
1 MJE15032 T8
1 MJE15033 T1
2 SK104 KK1, KK2
The BOM is an Eagle file, I pasted it into 'text' format for anyone who doesn't have Eagle.
Qty Value Parts
4 X1, X2, X3, X4
4 1µF C6, C12, C14, C25
12 1N4004 D1, D2, D3, D4, D5, D6, D11, D12, D13, D14, D15, D16
10 1k R1, R2, R4, R5, R10, R12, R19, R20, R21, R22
4 3k3 R6, R11, R13, R17
2 10V D8, D10
4 10k R3, R7, R16, R18
2 10k P1, P2
2 22k R8, R15
2 100R R9, R14
12 100nF C1, C2, C3, C8, C9, C13, C15, C16, C17, C21, C22, C26
2 100pF C7, C20
4 470uF 25V C10, C11, C18, C19
4 1200µF 35V C4, C5, C23, C24
3 BC546C T3, T4, T7
3 BC556C T2, T5, T6
2 GREEN 5mm LED1, LED2
2 LM4040D10 D7, D9
1 MJE15032 T8
1 MJE15033 T1
2 SK104 KK1, KK2
Scoring is a largely manual process that adds ~$.40 to the unit cost.
If the boards are scored they would need support in the center, even if not separated due to the weight of the parts.
If there is enough interest we could have the boards scored.
Jimbo, you'd need to have the large heat sinks, with some air circulating around them for your 5V supplies. (and either the 2 or 2.5V LM4040. Stacking wouldn't be a good idea. It would also be quite tall with 2" high heat sinks.
If the boards are scored they would need support in the center, even if not separated due to the weight of the parts.
If there is enough interest we could have the boards scored.
Jimbo, you'd need to have the large heat sinks, with some air circulating around them for your 5V supplies. (and either the 2 or 2.5V LM4040. Stacking wouldn't be a good idea. It would also be quite tall with 2" high heat sinks.
Since the PSU layout now uses the same size output caps as the filter boards' inputs, why not buy them all together?
Panasonic FC 470 uf/25V 10 mm dia x 5 mm spacing 10 for $2.80
And since everyone needs 100 nf bypass caps,
Epcos 100 nf MKT 63V 2.5mm x 7.5 mm 10 for $1.00
Sign up on the WIKI/email me with your requirements.
Bob
Panasonic FC 470 uf/25V 10 mm dia x 5 mm spacing 10 for $2.80
And since everyone needs 100 nf bypass caps,
Epcos 100 nf MKT 63V 2.5mm x 7.5 mm 10 for $1.00
Sign up on the WIKI/email me with your requirements.
Bob
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.