I am using a wallwart AC-AC transformer rated at 12V AC (measured ~15V) so exactly what I need to get +/-12V DC with those regulators without wasting much on heat.
Hook it u up to AC1 and GND and youāll be good.
1 Like
I just did, but it did not work still. No LED indicator or any voltage out 
Then itās time to start poking around with the multimeter. You should have between 16 and 17V on the big caps. Are the trimmers set correctly?
Shouldnāt I have the LED on, or something at the output with the trimmers halfway through?
Unless you have set the output voltage above the supply voltage, I guess. Check the voltages before and after the rectifyer
2 Likes
I got it to work. Dodgy barrel connector! Thanks!
5 Likes
Heya THOGRE your mofosāish psu w/bus board rocks dood, im running 11 modules on it and using bout 75%'ish of the max power cap.
i be bandwidth bussin on a budget!
1 Like
I have a āWhat would you do?ā question for a frequency central routemaster pcb i wanna finish rite here and it calls for 2K cermet trimmer but i aint got dat, i only have 2k variable trimmers.
i understand this trimmer is to adjust the voltage 12-15v and cermet is recommended here over a carbon composition trimmer?
Am i about to summon the magic smoke genie?
Unfortunately FC doesnāt seem to publish a schematic, so itās hard to be certain whatās going on, but I presume the LM337 is hooked up similarly to the datasheet, with the trimmer and the 1k being the voltage divider that controls the regulator. I donāt see anything in the build document that specifies cermet, just āmulti turnā, though there is a link to a cermet trimmer from Tayda. I also donāt see any recommendations in the datasheet. The current through the trimmer will be 12 V / (1k+Rtrim) where Rtrim is 0 to 2k, so 4 to 12 mA. At the midpoint the current is 8 mA and power dissipated is I^2R = (8 mA)^2(1k) = 64 mW. If thatās well within the specs of whatever trimmer you have Iād think it should be okay. And if itās not multi turn that just means itāll be touchier to adjust, but you donāt need high precision anyway. Iād just make sure to start at maximum resistance and work down from there. And keep sniffingā¦
1 Like
Yeah u rite Onward and upward!
The 25 turn trimmer makes sense to fine tune the voltage vs the oneway or the otherway variable trimmer. i frickin love this site, Thanks my friend
Please read my Routemaster saga here: DIY Power Supply Thread - Questions and Help - #225 by TimMJN
Check your board against my images, see if it was updated. If it wasnāt, itās a shit design.
6 Likes
Good eye tim!
i have the Rev 3 from 7/22 and the traces appear the same just a fast once over. thanks for documenting otherwise i would have been trippin
Edit: After assembling the FS Routemaster i proceeded to set the +12/-12v per the onboard trimmers , first set the -12 side then the +12 but the +12v led indicator dimmed and the voltage dropped just below +5v. I put the meter on the lm317 output pin and that was under +5v. Checked the rest of the board for component placement, bridges etc . Replaced the lm317 and no bueno. Trippin.
TimMJN is this what you encountered? I read your previous posts concerning the ground placements but donāt have a scope for further exams
Iāll dig deeper into this pita this weekend.
2 Likes
I was looking at the difference between using an AC transformer with a centre tap vs one without and have been scratching my head, trying to see what is the most efficient approach (current-wise but also cost-wise). Perhaps other members of this forum can advise on the issue. Here is the puzzle:
Suppose we have a 12V AC transformer rated at 3A.
A power supply design with a half-bridge rectifier (Frequency Central, MFOS, Analog Output, AI Synthesis, etc) has an efficiency of about 40% at the output, but assuming that we use the LM7912 regulator, we can get this 40% on each rail, so (unless I got this wrong) we can draw something like:
1,2A at the +12V output, and
1,2A at the -12V output.
Using a 3A AC transformer with centre tap, we can achieve an efficiency of 80%, but apparently (unless, again, I am wrong), this is shared between the rails so we end up again at:
1,2A at the +12V output, and
1,2A at the -12V output.
If the above is correct (and I think that it probably isnāt), then what is the point of using a transformer with a centre tap and a full-bridge design?
With a dual transform, the AC peaks reach the regulator at twice the frequency, reducing the ripple
I think the way to think about this is, for a 12 VAC supply and Ā±12 VDC output, a linear regulator has an efficiency of about 80% (Vin/Vout, where Vin is the rectified and filtered 12 VAC so is around 15 V). For a half wave rectifier design each railās rectifier passes only half the input (positive or negative half), so you end up with 40% efficiency ā essentially you have about 1.5 A after each rectifier and after the regulator you have (optimistically) about 1.2 A.
For a full wave rectifier dual supply design the rectifier passes all its input, but the transformer sends only half its output to one railās rectifier and half to the other, so again you have about 1.5 A after each rectifier and about 1.2 A after the regulator.
If you were talking about a single +12 V supply then yes ā a half wave design discards half the input, full wave keeps it all, and then itād be 40% vs. 80%. But for a dual supply with center tap transformer, since the full wave design uses only half the transformer output for each rail, it comes down to 40% vs. 40%. Thatās assuming Iām not confused, not always a good assumption.
As @TimMJN says, the difference is lower ripple with a full wave design.
Thanks @analogoutput and @TimMJN for the responses! It seems then that I got this right. It does not make sense to use a transformer with a centre tap for efficiency reasons for a +/-12V output, given that transformers with a centre tap are typically more expensive and difficult to find than those without. No wonder why I havenāt seen this as much in synth DIY.
There is the question of ripple of course. In the designs that I am comparing, to address ripple, the half-wave design uses three 4,700Ī¼F capacitors before each regulator, while the full-wave uses just one 1,000Ī¼F capacitor. For the full-wave the ripple factor is ~0.5 and for the half-wave ~1.2. So given that the capacitance in the half-way design is ~14 times that of the full-wave, canāt we conclude that the ripple is less in the half-wave design, in this particular comparison?
Ripple isnāt at all linear in the capacitance, and in fact Iāve seen it argued that 15,000 ĀµF is too large, though I donāt know enough to say whether thatās true or not.
I think I disagree here, we shouldnāt confuse different efficiency definitions. During the āactiveā phase of the rectifyer, efficiency (energy out/energy in) is 80%. During the āinactiveā phase, no current is drawn into that rail, so energy in=0. So energy efficiency remains 80%.
Good day fellow enthusiasts and wiz kids
My Frequency Central Routemaster has been giving me the blues lately.
I got the pcb a while ago and been putting it together as time allows. I read up on what our Tim here went thru with noise and issues with the lm317 and I thought I might have the same problem.
So I added a jumper to shorten the ground path hoping that would help but I donāt have a scope to scope any further.
Whatās happening is the +12v & +5v rails are good but the -12v side is not
At the lm317 input im reading -17.5v and -2.5v @output , the output voltage changes with the trim pot from -2.5 to -1.8v.
Iāve replace the lm317 three times and the diodes, resistors and caps I circled in this picture around that part of the circuit
I looked at all the traces and components values and Iām missing something obviously
Iām up for any suggestions at this point before trying
a different psu, itās been in the āstackā for too long
Happy Halloween
Edit: Iām using a Jameco ReliaPro 12vac 1500ma