FY6600 Function Generator Power Supply investigation and fix
Вставка
- Опубліковано 5 жов 2024
- The Feeltech FY6600 is a very useful function generator but let down by its cheap switch mode power supply, which like most power supplies of this kind, puts stray AC voltages in the 100V range on the output terminals. Its not dangerous - you may only feel a tingle - but it can upset sensitive measurement equipment as happened to me.
This video is about my investigation into ways fixing this power supply issue. In the end I used a very inexpensive way that got rid of most, but not all of the stray voltages (from 100V down to 1V).
A better fix will require more design work because the second problem of the FY6600 - its tendency to get very hot due to the inadequate ventilation of the enclosure. I'll post a video once I find a solution for that.
EDITED: 21-Nov-2018: watch the follow-on video • FY6600 Function Genera...
replacing the complete power supply
The video where I noticed the stray voltages first due to the effects on the Solartron 7150+ is here: • Solartron 7150+ part 3...
The parts list of the items used in my modification is
1x Schurter C6 Panel Mount IEC Plug Male, 2.5A, 250 V ac RS-Components Stock no.:498-2899
1x Power-cord C5 to UK-Mains plug (I had one available)
2x KEMET Paper Capacitor 4.7nF 480V ac ±20% Tolerance PME295 Through Hole +115°C RS-Component Stock no.:533-0609
Hello and thank you for this video. My unit became quite erratic after a year and then not long after failed to start up. It would pass it's so called self test but the channels would not start or produce any meaningful output. Bin it and buy a new and better model! No I decided to have a good look at it and thanks to your video I investigated the power supply. +5V, check, -5V, check, +12V-err 18V, -12V- err -16V. Right whipped out the two 220uF 16V electrolytics and replaced with 220uF 24V jobs. Result. +5v, -5V, +15V, -15V. The box now boots and functions perfectly, well to the best of it's ability!
Thanks so much for this incredibly thoughtful and detailed video! Every FY6600 owner needs to see this!
Best is to put grounding connector on the front panel. That way the device can be grounded however is suitable for the particular situation. Some setups require isolated outputs.
Yes, exactly. By the way, there is a similar issue with the USB port of this generator. If you have this permanently connected to a PC for remote control as I have, you need to use a USB isolator or the generator output will be grounded through the USB cable. A very nasty surprise if you were counting on it being floating... thank goodness there was a quick fuse in the circuits power supply that it was connected to...
Regarding option 4. 'TDC Power Products Co' make a suitable cased transformer 240VAC Secondary @ 80mA [19VA] for illuminated bathroom mirrors (for the shaving socket) Model # DC-240-0080T. These can be purchased in the UK from online bathroom mirror suppliers from their 'spares' pages. Typically about £16. Mine I got for free, because I salvaged it from a heated / illuminated bathroom mirror I replaced some time ago when the mirror rear silver failed around the edges. For mine, the dimensions excluding mounting tabs is 53mm long by 50mm wide by 43mm high. Weight 471 grams. Easily small enough to fit within the existing free space.
Heinz, excellent! I love your thorough consideration of a problem and your creative solutions. Thanks for your effort to share with us!
Thank you for your thoughts and work on coming up with a pratical and low cost fix on this issue. Along with your Y capacitor rework, I am planning on adding a 26VA ( 220ma @ 120VAC secondary) line isolation transformer that will cost about $15 USD to purchase to add a greater amount of safety. I will also consider opening up the molded in vent slats in the bottom of the cabinet, as well as possibly using a dremel tool and cutting some slots into the top panel molded sides to help remove heat.
Thanks for the nice words. For the price, the FY6600 is a great little function gen that can do amazing things. I have now started to use it mostly remote controlled, like doing Bode diagrams with the FY6600 providing the input and the output measured by my 34401A captured in a spreadsheet. Works great. Just be aware that if you connect the USB port to a PC the FY6600 output becomes automatically ground referenced. To avoid that, I added a bi-directional USB isolatator (uses optocouplers)
@@TheHWcave - Hi again - I am planning on leaving the 6600 "ground" floating, because I will mostly be using it with a connection to a oscilloscope , which will drain off most any stray voltage off of the unit ( should be very little to none after using a step down-isolating transformer) but I will add a fuse ( probably about 1/4A) to the primary/line side in case of any catastrophic device failure that may happen while using this unit or connecting it to another. I am also planning on using this for bode plots for alignments, but will need to add a good selectable attenuator unit in path.
This is how it should be done! A true professional approach.
Damn, you're absolutely right. I got 95 to 105 volts depending on whether I was standing barefoot on the floor or on a piece of carpeting I use for my feet while working to avoid being grounded. My Feeltech came from China to Thailand with a European type plug with no ground pin.
Good solution thanks, also if there was an electrostatic shield in the switching transformer it could solve this problem. I once replaced a switch mode psu with the once manufactured my mean well and they’re very well for that matter.
Very nice video! Excellent explaining. Very cool that you show your reasoning by presenting different avenues to a solution. IMHO you also have the right tempo for the audience.
That was a fantastic insight. I am definitely going to to perform this modification to mine when it arrives in a week and a half.
Well done, indeed.
That 'Y' cap. between primary and secondary in a SMPS that gives you 80 or 90 vac on the output ground _really_ annoys me :o) I know it's there to reduce RFI, but as you've shown, there are better ways to do it - the bad way should frankly be outlawed ! Thanks for your in-depth analysis of the possible fixes in your case - I think precision lab test gear _really_ should have a much better PSU than this - though I know the heat from a linear reg. and tx. might be a worse problem in some ways :o)
O boy. Ty for video. I changed my mind and will not make mistake and buy this half product.
I think they made some improvements. For example, I believe the power plug is now 3-pin (i.e. with earth) which gets rid of the phantom voltages but (I think) makes the output ground referenced which is not always what you want.
thank you very much, useful video
I was told that the lack of cooling slots was to obtain thermal and frequency stability.....but probably not for long.
I did this mod in February 2018 and dropped from 33.2v to 31mV not bad, eh? However, I used two 10nF.
Great video .... I'm thinking I might just make an external linear supply capable of running the Generator , A shielded multi cable would link the supply to the generator , Possibly the standard 4 pin screw lock plug and socket arrangement ( Same plug and socket as a lot of Ham radios use for the microphone / Mic socket would easily handle the voltage and current ) .... This would free up the inside of the generator to allow a fan to be installed in the rear ... 🤔🤔🤔 .. regards de Gw0wvl 👍
Some models come with the so-called wall wart, which appears to be a better choice.
Excellent content. Would share a lab with you any day! Thank you.
Is this why i get really bad hum on my amplifier circuits when running it from a feeltech? If i touch the bnc the hum decreases. It also seems to make my osxilloscope unhappy at times and i get a strange distortion signal, whuch takes ten minutes to go away on the scope ( it looks like triangle wave buzz)
Excellent !
I have modded my FY6600 with a RD-3513 Mean Well Dual Output Switching Supply. I used a step down buck converter (12 V to 5V). I replaced the oscillator with a Connor-Winfield (VA) / D75J-050.0M. I replaced the existing opamp with two Texas Instruments [CI] / THS3095D. There are a number of variable resistors used to calibrate the opamp offsets. I have noticed the FY6600 seems to loose the calibration over time and generates Vpp of 12V instead of 5v, on both channels, needing recalibrating. Any ideas.
Do you think the isolation transformer fitted within bathroom shaver sockets in the UK would be suitable for using in the PSU I understand the rating
for these has to be between 20 min and 50 max VA. These units can be obtained at quite reasonable prices.
I never looked at those before and yes, if they are well made (I mean isolation-wise) this is a possibility. 20VA should be good enough. One issue, looking at the regulations for these things, it appears that the max. allowed output voltage under load is 250V and voltage at no load is not to exceed 275V. This is worryingly high. I wondered if the original SMPS can handle it. I dug it out from the spare parts bin and the rectifier diodes (1N4007) are good for 1000V. The main electrolytic cap is rated for 450V, so you should be ok.
Thanks for raising this. I am now thinking of getting one just as a (low cost / low power) isolation transformer for the bench.
hi, this is a v good vid, i like it and you have payed alot of attention to it. this queation is a change in subject but, could you tell me please what the peak to peak output is from the fy6600 for a sine wave at 28.5mhz, 30mhz and 36mhz into a 50ohm load. thanks
28.5 MHz : 2.44Vpp 30 MHz : 2.42 Vpp 36 MHz: 2.38 Vpp all using a Rigol DS1054 built-in measurement function.
great job
Very interesting video!. I am following the mod videos on the FY6x00 topic for some time now and came here through a link provided by Tony Albus. Since I have a FY6200, which is the panel version fed by a 5V 2A SMPS, I thought I might check the stray voltage as wel. I found 72VAC measuring the same way as you did! Where does this come from? What struck me even more was when I plugged the wall PSU in my isolation transformer, the stray voltage was still there! You demoed an isolation transformer even before fixing the problem. My non-contact AC tester goes crazy as well, showing AC (green) in the vicinity and AC (red) when very close (which corresponds with the measured values. How can I still get these high stray voltages even when using the isolation transformer?
I was wondering whether you could shed any light on this problem and give my any tips for a possible solution. Thank you!
Hi Marcel, in a standard SMPS the stray voltage comes from the Y-capacitor that "bridges" the high-voltage and low-voltage sides. It is there to reduce switching noise on the output. A reasonable large (250VA) isolation transformer 230V - 230V (I have a proper one now) has a considerable capacitance between the large primary and secondary winding. Mine even has a grounded layer between primary and secondary and I still measured 0.31nF beween primary and secondary. That isn't much but sufficient to cause stray voltages. Interestingly, smaller transformers, I mean normal transformers like 230V primary and 2x15V secondary have hardly any capacitance, 2-4 pF or so (I used my normal multimeter capacitance mode for this, not a proper LCR meter). It kind of makes sense because these transformers are physical smaller and the secondary winding is not that big either because the output voltage is relative low. Fact is I have an old (20 years+) traditional very simple linear power supply using one of these transformers which does +5V, +15V and -15V using 78xx and 79xx regulators on heat sinks and that has only 3V stray voltage. My Fluke voltage tester which is really sensitive to leaks stays silent on any of these outputs. This matches my experience with the modified FY6600 power supply that I still use which also uses a small traditional transformer.
What can be done? For one there are SMPS that are certified for medical use. I have used some of these (from Mean Well, but others are available) and they had almost no leakage. I think they do not use Y-caps as normal SMPS and must do more effort in filtering the output, so more expensive. I did not know those existed when I did my FY6600 mod. I am not sure how far these solve the problem but it may be something to look into. If you only need 5V, they are easy to find. Medically approved SMPS that provide 5V plus additional voltages are much harder to come by. The second option is what Feeltech has apparently done for the FY6900. Simply connect the ground of the SMPS output to earth. To do this they have apparently changed the power plug at the rear of the generator from 2 to 3-pin and now supply a 3-wire power supply cord with earth. That works but at the cost of referencing your generator output signal to mains earth, meaning you have to apply the same caution as when using a mains powered scope. See Dave Jones EEVBlog "how not to blow up your scope" video if you have not done so.
@@TheHWcave Hi, thank you for the comprehensive answer :-) I've been experimenting a bit now. Using serveral smps psu units (AC-DC or AC-DC followed with DC-DC buck converter) I cannot get lower stray voltage then around 10-15V. Since the FY6200 only uses 5V DC, I think I'm going to make a small linear (7805) psu with stuff from the bins. II'll breadboard it up first and see what it will do.
Before watching the videos on stray voltage I was not aware of this. What measurements would influence this most? When showing on the scope I really don't see any effect. Maybe the RF domain is most effected?
@@TheHWcave Forgot to mention a few things :-/ i almost take it for granted that you've also watched SDG Electronics videos on this particular topic, the power supply for the FY6900? I watched them again tonight where he designs and builds a completely new power supply. Very interesting videos. He takes a ready made AC-DC module, then 3 smps converters and the plus and minus 13.5V rails are following by two linear regulators. Found that an interesting concept to remove switching noise. Then for the floating AC voltage he just put a 100ohm resistor between the Ground of the 24V output (or regulator pcb input) and the earth ground. Of course, you make the ground connection earth referenced just as your scope, but the AC was completely eliminated.
I also know about Dave's video :-;
What effects stray voltages have is hard to say. It depends a lot on the test setup, what are you trying to measure and how sensitive is the device under test. So your mileage may vary as they say. In any case its useful to be aware of that stray voltages could be around.
Your approach sounds sensible, but a 7805 is maxed out at 1.5A even with excellent cooling. So you need to look out for this. As I explained in the videos the problem is the rectified DC us usually high which causes a lot of heat in the regulator and when you design the rectified DC to be "close" to the target, you run the risk of ripple getting through. Not easy. ..
@@TheHWcave I learned a little something because of this. I don't really have to worry about this little power supply. I might even have a different solution. It's funny that this FY6200 has the same functionality (as far as I can see) as the FY6900 and only runs on a single 5V supply. It does only consume 850 mA max when running both channels with a 30MHz multitone wave.
Hi interesting video there, I saw some 30w isolating transformers on Aliexpress can I use one of those with the existing smps and not fry my chips without any other additional circuitry?
You would still get some stray voltages but they would be isolated from mains, so that's good. Just make sure that transformer is of good quality and really keeping the two windings separate, preferably they should each be in their own isolated compartment side by side and definitely not wound on-top of each other. Have a look at the Wikipedia article on "Isolation transformer" en.wikipedia.org/wiki/Isolation_transformer
Hi,I got a 30va transformer with windings isolated from each other and wired it in.The stray voltage fell from 119volts to 11,6volts.I hope this won't fry my fet opamps.I
That thing is deadly, holding the - on my voltmeter I get 128 vac on the pnc connectors.
Hi Robin, it is high voltage but very low current (just the leakage through the Y-capacitor), so it wont kill but you might feel a tingle or roughness. The main problem is that this leakage screws up sensitive measurements
@@TheHWcave Thank you for the answer, Aliexpress has some small 220 in 220 out isolation transformers in the $20-30 range. www.aliexpress.com/item/33023544948.html?spm=a2g0s.9042311.0.0.2ea64c4dkzRmz8 How many watts/amps of an isolation transformer would I need on the primary side (option 4)? Thank you.
It's very humid here in Thailand and I like to work with bare feet which makes the problem worse.
I know what you mean, I lived in Singapore for quite some years. Anyway, I looked at my records and based on the currents I measured from the original power supply, the combined DC load of the function gen. PCB was around 6 Watts. Assuming the SMPS manages 80% efficiency, it would draw 7.5W from the mains. Given the uncertainty of this estimate, I would go for at least 15W or 20W for the transformer. Hope that helps.
Brasil 👀🖒
Adding an external power supply with a 4 pin plug & socket would keep the heat away from the unit
Totally agree with you, although its a bit of a pain to have a 2nd box (external PS) and you don't want the DC cables too long.
I have had a problem with the KK Moon variant of the same generator when it is sat next to other instruments on the bench, the only way I could stop the jitter (at any frequency) was to raise the back of the device so it was not at right angles to the other equipment. It suffers from EMC, my device is not ce marked but a good instrument all the same.@@TheHWcave
Everything from China seems to overheat, they must be working in conditions of snow and ice and are using the heat from this equipment to stay warm. You are greatly overthinking this, I would just drill some ventilation holes in the top, remove the blue cap, ground the outputs and let it got at that.