Schematic discrepancies

[Kristofer Danner]

Populating an end plate to add a tuner and fan here. 

According to this: https://github.com/softerhardware/Hermes-Lite2/wiki/IO DB1 pin 4 is FAN, pin 5 is ATU ACK, and pin 6 is ATU RQST.

I connected according to this, more on this in a minute.

But here: https://github.com/softerhardware/Hermes-Lite2/blob/master/hardware/companions/io/io.pdf pin4 is ATU REQ, 5 is ATU ACK, and 6 is FAN. So 4 & 6 are swapped vs above.

Now I don't find anything actually saying I need to connect the DB1 FAN pin to a J2 FAN pin. Do I or not? Because when I connected it, it smoked the fan.

Another thing is I've got 2.2V on pin7 of the terminal block for the ATU, not 5V, but again, at this point I'm not even sure I have the wires on the right pins due to the discrepancy above.

 https://github.com/softerhardware/Hermes-Lite2/tree/master/hardware/enclosure/endcaps/kf7o/hl2_55b Shows J2 pin 8 to be CH4 but I think it is actually CH5.

So if someone could tell me which document is actually correct, I'll try to follow that one and troubleshoot from there.

Thanks,
KD

[vk4...@gmail.com]

I only use the fan. I connect to DB1 pin 4 as in the wiki/IO and this drives Q1 as in the io.pdf. The pdf showing connecting to pin 6 appears to be wrong. The connection from DB1 pin 4 goes to the FAN on J2. I soldered mine at the J2 end.

73 Brian.

[Steve Haynal]

Hi KD,

The wiki is correct and the pdf schematic is wrong. In fact, io.pdf was an experiment I drew up which was never built. I should remove it from github. The latest and current end plate information is at:

https://github.com/softerhardware/Hermes-Lite2/tree/master/hardware/enclosure/endcaps/kf7o/hl2_40b

You need to connect DB1 FAN pin to J2 FAN pin. The fan does not connect to J2 directly but through a transistor switch as described here:

https://github.com/softerhardware/Hermes-Lite2/tree/master/hardware/enclosure/endcaps/kf7o/hl2_40b#fan-active-components

The schematics here:

https://github.com/softerhardware/Hermes-Lite2/blob/master/hardware/enclosure/endcaps/kf7o/hl2_40b/endcap.pdf

Indicate that the fan operates at Vsup, usually 11V-16V. Was your fan rated for that voltage?

Yes, CH4 appears twice in the picture at the link below. The CH4 between ISOIN and FAN should be CH5.

What ATU are you using? Do you have level shifters installed for the ATU?

73,

Steve

kf7o

[Kristofer Danner]

Hi Steve, well that explains why the fan smoked. I used a 5V fan because that is what is shown in the photo at https://github.com/softerhardware/Hermes-Lite2/tree/master/hardware/enclosure/endcaps/kf7o/hl2_40b

The ATU is an MFJ-998. I would not normally purchase anything from them, but I acquired it along with a ton of other ham gear for free recently. It claims to be AH-4 compatible so I thought I would give it a shot. I did install Q4B and Q6B along with the associated resistors as indicated on the wiki here: https://github.com/softerhardware/Hermes-Lite2/tree/master/hardware/enclosure/endcaps/kf7o/hl2_40b#atu I guess my next step will be to remove all of these jumper wires and just connect the minimum for the ATU to see what I have then, go from there. 

It's getting to the point that there are too many of these individual wires so I was wanting to get things like 16 position connector shells, make up a harness which can be removed when needed, yet keeps all the wires in the right place. I also wanted to put a 12 position header on the DB-9 anyway, for the pullup resistor to connect to. So I guess I'll order all that while ordering 12V fans. To do it up really nice probably the thing to do would be to use shrouded headers with keying slots so that the mating connector could only be connected in the proper orientation. When I get a 2nd HL2 that is probably what I will do to it, but at this point I don't want to desolder DB1 plus all the headers I've already added to this one.

I'd like to help with clearing up some of this on the wiki, but I fear I'd have to pepper you with a lot of questions. When I see things in disagreement with each other I just often don't know which one is correct.

Thanks,
KD

[Kristofer Danner]

Alright I believe these connections for the ATU to be correct. I still only see 2.2V on ACC1-7, not 5V. Also if I remove all add on wires and check voltage at DB1-5 it is 0.5V, not 2.5V. All other voltages are correct per the wiki. Suggestions?

Thanks,

KD

[Steve Haynal]

Hi KD,

Regarding the 5V fan in the picture, I can't remember if I was just lucky that it worked at 12V or if I was playing with a lower voltage. The picture is misleading. I did buy about half a dozen 12V fans of the correct size from Ali Express to experiment with. I can send you one in a padded envelope if you are in the US.

I was not happy with the results from the fan. There is data somewhere on this list, but IIRC the fan only lowered the temperature by a few degrees, even at the hottest points of operation. I think the fan is just too small for the job. Others have had success with pointing a larger fan at the unit. The heat shim is another better option. I personally think people are often too worried about the temperature, one of the problems created by reporting the temperature. It is okay if the unit reaches the low 50Cs during operation.

I think the tuner has only been used by a couple of people with a true Icom AH-4. There may be issues with compatible AH-4 tuners.

The original idea with the end plate was a lot of flexibility. But I agree with you that this flexibility has often led to a confusing bunch of wires. Maybe it is time for an end plate 2.0 with just one ribbon connector from DB1 plus the Vsup and GND.

You wiring in the picture looks correct. You shouldn't expect a 5V swing unless there is something connected to the ACC which pulls it up. Once the AH-4 is connected, you should see the larger voltage swings. This is just the nature of the level shifter used here.

One last note, you may want to run some extra ground connections. This has helped with noise for noise user.

73,

Steve

kf7o

[Kristofer Danner]

Hi Steve. Thanks for offer on the fan but I ordered this one already, 6CFM and only 27dBA. https://www.mouser.com/ProductDetail/369-MF30151V11UA99

My goal is to be overkill on the cooling. I started with the 40mm enclosure and shim. If band hopping a few bands with WSPR where I was transmitting every 2min cycle, I would see 55C by about 12 minutes. External fan over the enclosure did care of this and kept it down around 40C, so I'd rather clean things up with an internal fan. When I switched to the 55mm enclosure I did put a 3.5mm thermal pad in like the shim in the 40mm housing. Not because I expected it to be any better, but just to make it as good as I could. This also takes about 12 minutes to reach 55C. I was a bit surprised that it seems to work just as well. Hopefully the fan will get it down to similar temps seen with the external fan.

I just tried the ATU and you are correct I do see the 5V when it is connected. It works sometimes, on some bands. The thing is, it seems to work(or not work) just the same off RF sense if I disconnect the cable to the HL2 so it may just be pointless to have a radio interface to this thing. I'm sure part of the failures is it is a QRO tuner not seeing enough power, sometimes when it fails to tune it will even say increase power on the display. I will try it again when I have an amplifier finished, but I will never have much expectation for anything from MFJ.

I did also order some wire housings to fit all the headers so hopefully in a couple of days I will have that cleaned up and a working fan.

Thanks,

KD

[Kristofer Danner]

Steve, I have a working fan now and it is getting the job done. It took a full hour of nonstop transmission into a dummy load to get to 52C. Likely to be a bit less in real world operation with the few seconds break at the end of each 2min WSPR cycle. It is probably good enough but I think I will still attempt to improve it some. I removed the front end plate to see if it might be restricting airflow but this made no difference. I wonder if the rear end plate might be restrictive so I may just sacrifice one by cutting it out into one big fan hole.  Another possibility is there is a 7.7CFM 30mm fan available and/or 3D printing an internal fan duct so that the air is drawn in close proximity to the pair of AFT05MS003.

KD

[ron.ni...@gmail.com]

I am wondering whether blowing air on the outside of the case, just above and below the slot where the PCB is inserted, can remove more heat from the edge of the PCB than a fan inside the case, which can only remove heat from top side of the PCB.  I have a small Raspberry Pi heatsink-fan combination, but I haven't yet figured out a good way to mount it outside the 40mm case, but against the internal slot area.
73,

Ron

n6ywu

[Kristofer Danner]

Made some progress on this today. I crudely cut out this endcap then populated it for the fan. After the same 1 hour test with the front endcap off, temp was 48C. Retesting with the front endcap on was 49C. So it looks like I am going to be learning to edit gerbers, get some boards made with the big hole in the rear and more ventilation holes in the front. I'll get the 7.7CFM fan coming too.

KD

[Steve Haynal]

Hi KD,

Glad to hear you are achieving some good results with a fan. One other idea which has come up is to mill a proper opening on the top of the enclosure and then use a larger fan to blow directly down on the PCB.

73,

[Kristofer Danner]

Hi Steve, yes I've considered similar. Some blower type fans could work like that and move significant air. Another thought I had was it looks like a 36mm fan can fit in the front endcap and some of those can move over 3 times the air that any 30mm can. The problem is they are quite loud. Anyone that has spent any time in data centers with a lot of 1U servers will know the sound, just high pitched and annoying. Hoping to avoid anything like that and I think the 30mm in the original location is going to work quite well with just a few adjustments.

KD

[Kristofer Danner]

New endcaps with 7.7CFM fan, plus wiring cleaned up. I believe this is as good as I can do without modifying the enclosure, and it is certainly good enough.

Ambient 26C

Quisk showing PA at 1600ma 

1hr continuous TX - 47C

1hr WSPR TX every 2min cycle - 46C max, drops to 41C before beginning next TX

1hr FT8 TX every other 15 second cycle - 41C max, drops to 36C before beginning next TX

KD

[Steve Haynal]

Hi KD,

Looks very nice. You can add your new end plates to the github repository if you like. Just issue a pull request.

73,

Steve 

kf7o

[Kristofer Danner]

Steve,

 I'll do that. I did send a pull request a couple of weeks ago for some other things. Looks like you may have missed it.

Thanks,

KD

[Steve Haynal]

Hi KD,

Thanks for the earlier pull request. Sorry that I missed it. I have merged it now.

73,

Steve

kf7o