Products > Test Equipment
FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
Noy:
Found this:
https://github.com/DerKammi/FY6600-15-30-50-60M/blob/master/Hardware/FY6600%20PS%20schematic%20V1.pdf
I think i will try to "upgrade" the supply If the ths3491 isn't enough.
Any suggestions for another/better pincompatible DAC?
DAC904 in fy6600 regarding the Link but dac904 are 165Msps Not Like Advertisement 250Msps. In my FY6800 are two AD8009 instead of the OPA686. Also any better suggestions? AD8000?
Yansi:
I think that changing these variables (DAC and OPAs) will not improve anything marginally, just a tiny bit at best.
The layout of the PCB is just horrid, that should be fixed first I think for any significant spec improvement to occur.
/* Just wanted to say Hi! Haven't watched this thread for a while (my taste for the FY6600 has gone) an am impressed by how far have you got so far */
Johnny B Good:
--- Quote from: Johnny B Good on February 05, 2019, 04:40:47 am ---
--- Quote from: Gege34 on January 31, 2019, 08:44:17 am ---Johnny B Good thank you very much for this great description to improve the FY6800 :-+.
I have the FY6800 from KMoon (firmware 1.7.1 but reported as 1.7 with the PC software).
1. Could you please make a picture of the internal modded device to see were you put the fan and the TCXO.
2. About the power supply upgrade, to summarize, what is the easiest way to improve the original one?
4. About the TCXO how to properly connect it (cable to use, ...) to the main board?
I have found the 50MHz version here (several of available).
Thank you
--- End quote ---
Sorry about the delay in responding. I had hoped to get the chance to take some photos to post here rather than having to say, as I now am, that I'll post some up as soon as I get another chance to open it up. Right now, I'm rather preoccupied with a little DIY GPSDO project to let me more accurately and swiftly calibrate the 50MHz 0.1ppm TXCO upgrade. As soon as I get the chance, I'll be only too happy to post the photos here and do a little showing off. :)
JBG
--- End quote ---
Well, after finding a remarkably easy fix for the extreme deafness of the on-board patch antenna on my "Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI" I bought from here: <https://tinyurl.com/y7ypyhnn> (ordered last Thursday, promised to arrive Tue or Wed, and in my hands the Saturday morning!), which simply consisted of a 14cm length (3λ/4 antenna) of wire poked into the SMA socket, I was able to justify popping the lid of the signal generator to tweak the TCXO module to within a couple of ppb of the 30MHz test signal I was beating against the third harmonic from the GPS module I'd programmed to output 10MHz on its PPS line. I'd already splashed out three quid on a mag mount active GPS antenna with 5 metre cable and the appropriate SMA male connector but I thought I'd have to wait on its delivery later this week before I could start locking onto any satellites, so it was an extremely pleasant surprise when my second attempt to use a simple wire antenna produced such copious quantities of useable satellite signals (10 to 14 decodable sats at any one time - I guess the magic charm was in using a 3λ/4 instead of a λ/4 antenna wire - who'd have thought that extra 10cms would be so important!). :)
Anyhow, buoyed up with such joy (and armed with a valid excuse to delve once more into my "Box of Tricks"), I remembered the request for some pictures and set up a photo-shoot. The picture quality could have been better but I was concerned about the file size limits in this EEVBlog forum. Still, I think they'll suffice for an initial look at what I've done to my FY6600.
9493 is a top down view
9499 looking over the front panel towards the rear panel
9501 Attenuator mod to convert from an 85 ohm to a 50 ohm impedance attenuator pad (it still needs more work to get it right though).
9502 XO location, right next to a source of 70 deg C heat (the three regulator chips to the left). The co-ax runs to the module to the left.
9503 The TCXO module mounted at a jaunty 45 deg angle above the fan to both act as a deflector and to remain the coolest item in the whole box so as to keep it within (just!) its 0 to 25 deg C range specified to meet its 0.1ppm rating and to shorten its warm up time.
9504 I needed to move the feet to both clear space to mount the fan but also, (and more importantly - requires the hard rubber feet to be replaced with something far more 'grippy' for best effect) to move them nearer to the back edge so that tilting it up on its tilt stand doesn't leave said feet dangling as the back edge of the case usurps the function of said feet.
You'll note the additional 8mm dia holes on the LHS of that image to supplement the existing vent slots and the opening up of the rectangular front panel clip retaining aperture just to the right of the leftmost foot near the front panel. You can't tell from that photo since I opened the rectangular aperture by bevelling its sides so as make such enlargement all but invisible to cursory examination. This helps reduce the heat that accumulates under the main board from the opamps's dissipation into the PCB (heat doesn't just choose to leave only via provided heatsinks. Heat's not fussy about how it escapes, It'll go via any and every route possible).
9506 and 9507 are side views to show how the issue with the badly placed rear feet was resolved.
9508 Shows the IEC C6 trefoil mains socket (less of the "Tail Wags Dog" effect risk that the thicker and stiffer C13/14 alternative entails).
9510 Is a shot showing the ducting over the rear vents to put them back into use as an effective contribution to exhausting warmed air instead of having to block them to stop freshly impounded air venting straight out the back unused. Its presence also assisted the deflection provided by the TCXO board in pushing the airflow towards the front from where it could traverse the main board in the region of the heatsinked opamps.
Regards to all, JBG
Gege34:
Thank you Johnny B good for the picture.
Where did you put the minus of the TCXO power supply TCXO (it's seems to go the the earth) and why there resistor (and value)?
Have you made any changes on the power supply (except the ferrite)?
Johnny B Good:
--- Quote from: Gege34 on February 06, 2019, 02:06:19 pm ---Thank you Johnny B good for the picture.
Where did you put the minus of the TCXO power supply TCXO (it's seems to go the the earth) and why there resistor (and value)?
Have you made any changes on the power supply (except the ferrite)?
--- End quote ---
Hi Gege34,
I used the co-ax braid for the ground return (-ve) to avoid ground loops. The white wire is the 5v feed picked up from the +5v pin on one of that group of regulator chips. The resistors you can see are a bunch of three 33K resistors in parallel using the -ve screw terminal on the module as a convenient connection point with the other end of the resistors wired to the safety earth pin on the mains socket (it knocks the 90v leakage right down to just under half a volt and neatly avoids mains earth loop issues.
As far as the smpsu board goes, I've lost count of the number of times I've had it out on the bench for the various modifications I've tried on it. However, the first mod was to wire a 47K across the lower half of the voltage feedback potentiometer formed by the two 10K resistors between the +5v and the 0v rail to boost the 5v to 5.5v in order to raise the 11.5 volts on the "12 volt" rails to somewhere around the 12.7v mark. This wasn't quite enough, even after replacing the weedy rectifier diodes with 20A dual shotky rectifier diodes so I ended up modifying the transformer to add an two extra turns on each end of the 24v centre tapped winding feeding the 12v rails. This raised the voltage to somewhere in the region of 13.7v or so.
In hindsight, I should have just upgraded all three rectifiers and then added a single turn overwinding to the transformer to buck the 5v winding, forcing the switching IC to compensate and so neatly increase the 12v rails without altering the 5v rail voltage and avoid modifying the transformer since there is ample room to thread (three or four lots in parallel of) the single turn of wire around the existing windings without having to remove the transformer from the PCB as I'd had to in order to to access the ends of the 12v windings in order to unsolder them from their solder tags so as to extend each winding another two turns each.
Connecting this one turn buck winding is just a matter of lifting the anode end of the 5v diode from the board and connecting the ends in series with the diode's anode and the vacated hole on the PCB. If you connect it series aiding on your first attempt (a matter of dumb luck), all that will happen is that the 12v lines will drop in voltage, leaving the 5v untouched which simply means you need to flip the one turn winding connections round to buck the 5v, forcing the 12v rails to increase in voltage.
[EDIT 2019-03-31]
I finally got round to undoing the original transformer mod yesterday afternoon to try out the single turn winding to buck the 5v secondary's output voltage as per the above. Not surprisingly, I had to reverse the connections to make it buck the 5v secondary.
It worked just as I'd hoped it would. However, when it was connected 'aiding' the 5v didn't get above 3.96v (I suspect maybe due to too large a smoothing cap on the 5v rail) leaving the 12v rails registering just over 8 volts each. Undismayed, I reversed the connections and tried again. getting +4.94v (I'd removed the 47K voltage boosting resistor), -12.98v and +12.83v. I wired in a 200K voltage boost resistor which then gave me +5.07, -13.29 and +13.13 volts which seemed to be close enough to the optimum. I didn't want to go much above the 5v mark in order to reduce the dissipation in those three very hot running LDOs on the main board and the 3.3v LDO on the 50MHz 100ppb TCXO oscillator board itself.
[END_EDIT]
Adding a single turn overwinding to the transformer (use three or four single turns in parallel for more efficient coupling), avoids leaving any evidence of alteration should you decide to remove it to restore it back to its pristine state and neatly raises the 12v rails without having to raise the 5v rail to achieve a similar result. In this case, it reduces power consumption and hence the waste heat produced by those ghastly regulator chips. The loss introduced by that single turn buck winding is so slight as to reduce the waste heat produced by the PSU itself simply by virtue of the reduced loading on the 5v rail.
As for the troublesome 'Touch Voltage' leakage introduced by the EMI bodge capacitor (the class Y 1 or 2.2 nF cap), the only effective remedy is to replace the IEC C8 two pin non polarised mains socket with a 3 pin earthed connector, preferably a C6 trefoil type rather than the C13/14 connector to keep the "Tail Wags Dog" effect to a minimum and link the protective/safety earth pin on the socket via a 10K resistor to the chassis ground rail. This will suppress the Touch Voltage without introducing a low impedance mains earth loop.
This, of course, only applies to the FY6600, the FY6800 is already equipped with a rather brutal C13/14 socket with a full galvanic connection to the earth pin socket bodged into the PSU to Mainboard ribbon cable connector, sacrificing one of the two zero volt rail connections as a result of this bodgery. FY6800 owners might want to fix Feeltech's earthing bodge if only to restore the zero volt rail connection back to its original integrity.
JBG
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version