I need some help with my FY6800 power supply upgrade. I did most of the upgrades shown in this schematic https://github.com/DerKammi/FY6600-15-30-50-60M/blob/master/Hardware/FY6600%20PS%20schematic%20V1.pdf. The only difference is I did not change D9 and instead of changing R6 I replaced the two 10K voltage divider with a 20K Multiturn Trim Pot wire the same as a voltage divider so I can adjust the output voltage. Also instead of using a 68uF 400V for C2 I used a 33uf 400V. Now the FY6800 intermittently turns on by having to turn it on and then off several times using the switch in the back. When it does turn on it works fine, with the power supply voltage at -13.6V, +13.5V, and +5.0V. and the on/off button in the front work fine as long as I do not turn it off with the back switch. Any help will be appreciated.
Thanks,
Mike
Hi Mike, and welcome to the FY66/68 hundred modders' house of fun and pains.
I had several criticism regarding those modifications when I looked them over for inspiration on 'improving' the PSU board in my then newly acquired FY6600-60M way back in November last year just before I made my very first post to this forum. I've inserted a link so you can see my initial modifications to the PSU board (I've lost count of the all the other mods I've tried with that PSU since then).
https://www.eevblog.com/forum/testgear/feeltech-fy6600-60mhz-2-ch-vco-function-arbitrary-waveform-signal-generator/msg1973777/#msg1973777 For starters, the value of the mains smoothing cap was over-specified by almost a factor of seven. The original 10μF 450v is entirely adequate for a 10W rated smpsu designed to operate on mains voltages in the 86 to 365 vac range (1μF per watt appears to be based on a universal design rule for all such universal mains voltage smpsus - I've yet to come across any exceptions to this 1μF per rated watt rule).
If you're only using it on 220v or 240v mains supplies (the harmonised 230v mains specification in all its guises of practical reality) without any thoughts of ever using it on 110/120v supplies, there's absolutely nothing to be gained by such an 'upgrade'. Indeed, the reduced voltage rating from 450 to 400 volt of the suggested replacement places the logic of replacing the 220μF 16v caps on the 12v rails with 25v rated 220μF caps in some doubt.
Whilst the 400v rating is perfectly fine for this task, the voltage rating upgrade on those 16v 220μF caps would equate to increasing a 400v smoothing cap to 625v (or the original 450v cap to a 700v rated one - it makes very little sense when the PSU is operating under its designed conditions (never disconnected from the load whilst powered up for prolonged periods of time) to uprate those output caps to 25v. A far more useful upgrade would be to replace them with 470μF 16v caps - the voltage isn't the issue here (it really isn't - trust me).
However, when it comes to improving the smoothing on the 5v supply, it can be all too easy to get carried away (replacing the 470μF with an 1800μF 6.3v cap and replacing the 220μF with a 1000μF 6.3v cap) and overload the dinky little high voltage switching IC on startup causing it to sulk in a shutdown standby state. In the end, I landed up reducing these capacitor upgrades in two stages before I could eliminate the hot restart issue that had appeared after I'd cured the cold startup issue. Effectively, I'd dialled back the 1800μF to a mere 1000μF and restored the original 220μF.
You could say (as indeed someone did) that I'd upgraded it to 'capacity'. If I hadn't already upgraded the 12v rail caps to 470μF ones, I might have gotten away with the 5v rail capacitor upgrade. However, using up my capacity margin on the 12v rails was the best option anyway - less ripple on the analog supplies (all derived off the 12v rails) where it really matters - the 5v rails can go hang, it's only used to power the three LDO regulators anyway (3.3, 2.5 and 1.8 volt rails) on the main board.
The diode upgrades are a definite requirement to both improve the 12v rail voltages and the efficiency of the PSU (apropos of which, install a small cooling fan if you don't want to be faced with a recapping job in 12 to 18 months time - the vent slots are counterproductive to passive cooling, a small vent fan is an absolute must imho). If you read that first posting of mine, you'll note I went a little overboard on the diode upgrade (well, I had them to hand, so why not already?
).
Much later on, I'd worked out an optimal solution to increasing the voltages on the 12v rails far more effectively than my 47k 10% boosting bodge on the 5v rail to likewise boost the 12v rails (over and above the extra two turns I'd added to each of the two 12v windings I'd applied in a previous modding session). ebel0410's mofification to the voltage feedback network raised a wry smile by the fact that he'd made more work for himself by
removing R6
and replacing it with a 12K rather than use the quicker and slightly more effective shunting of the other 10K in the voltage divider with a 47K to get virtually the exact same result. When it comes to this sort of modding work, there's usually more than one way to skin the cat.
You can view my description of my final PSU mod here:-
https://www.eevblog.com/forum/testgear/feeltech-fy6600-60mhz-2-ch-vco-function-arbitrary-waveform-signal-generator/msg2310768/#msg2310768 And, you might find this earlier posting of some interest too.
https://www.eevblog.com/forum/testgear/feeltech-fy6600-60mhz-2-ch-vco-function-arbitrary-waveform-signal-generator/msg2182397/#msg2182397 For all its faults (high ripple noise - on a par with any commodity smpsu to be fair), the one thing Feeltech's tiny smpsu board doesn't suffer from is imbalanced current output ratings on its +/-12v rails, unlike every single example of dual rail commodity smpsus I've ever seen on Ebay (whether just a naked PCB or an enclosed unit).
Looking on Ebay for dual rail smpsus designed to satisfy the requirements of dual rail opamps is on a par with tracking down Unicorn droppings (never mind the Unicorns themselves!
) Typically, such small dual rail smpsus will specify a +12v rated at 600mA whilst the -12v rail is only rated for 150 to 250mA at most, often with the requirement that the +12v be loaded to a minimum level to bring the -12v to within 10% of its nominal voltage spec.
Mind you, if you happen to be planning on adding a 12v 10MHz OCXO (with a 3N502 clock multiplier chip replacing the shite smd XO chip on the main board), such an imbalanced spec might be just the ticket!
Obviously, there are suitable smpsu modules available which have balanced ratings and very low switching noise but you won't find any on Ebay. You have to deal with the manufacturers or their agents directly and put in a request for a quote on pricing and you know what they say about that... "If you have to ask, then you probably can't afford it."
JBG