ct60 with a pico psu - startup problematic

Discuss CT60/CT63, CTPCI, SuperVidel and EtherNAT hardware here.

Moderators: Mug UK, moondog/.tSCc., [ProToS], lp, Moderator Team

tommo
Atari User
Atari User
Posts: 30
Joined: Mon Jan 29, 2018 6:00 pm

ct60 with a pico psu - startup problematic

Postby tommo » Sun Jan 20, 2019 7:01 pm

Hi all,

The ct60 works perfect, but i wanted a visual better solution. 8)

The big external psu has a switch whichs switches the falcon on/off.
When on it can now be switched on and off(sleep) with the power button (that connects pin 1&2.)
It works perfect, but i wanted a visual better solution so i got a pico.

Now the pico-psu does not switch on the falcon when powered.
If i get it to go ON it works fine and stays on. and the power button does indeed switch off the power.
but to GET it on is the problem.

When the switch is released the ct-power-led for a moment gets brighter(power on).
also the ven of the ct makes a turn and the green led of the keyboard (power on) goes on for a brief moment.
Rarly it starts up, but more likely after 2 quick pushes on the power button.
I read about the "Fix for CT60/63 that doesn't power on" 10uF on pin1 and 3 so i did that (not easy to do!), but it did not work.

I have tried the pico on 2 different 12v, both the same result.

This doesnot seem to be a "keep the power on"
but a "start the power on" problem. :shrug:

Forcing the pico with a switch that connect the green to ground works. (like regular on/off switch)

The pico is a 24pin just 34mm high 180w with a 10A 3.3v

anyone run agains this or has an idea?

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Sun Jan 20, 2019 8:11 pm

Nevermind.

stormy
Atari Super Hero
Atari Super Hero
Posts: 632
Joined: Tue Jan 26, 2016 12:39 pm

Re: ct60 with a pico psu - startup problematic

Postby stormy » Sun Jan 20, 2019 9:31 pm

Not sure what type of switch you are using, but you need a 'normally open' switch. So when you press it, it bridges the contact, then re-opens. Perhaps you are using a toggle switch or something.
Falcon ct60e 060 - 256mb ram - phantom bus and dsp accel // Atari TT - Thunder and Storm IDE 64mb ram - Lightning VME - USB LAN - Ati Mach64 2mb // Atari STM 4160 - 4MB ram - 16Mhz exxos 1.5 booster - Ultrasatan

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Mon Jan 21, 2019 9:26 am

I have meant to say off-spec picoPSU is to blame, specifically slow ramp-up of 3.3 rail, but I do not want to start flamewar.
But then I have realized it should be pretty easy to verify my thesis. If you want to do so:
Disconnect a picoPSU from power brick. Short ps-on pin to ground pin using a short wire. With CT60 set to 60 mode connect picoPSU to a power brick. Pico should start and if its on spec your Falcon will boot. If not, you will see fan spinning but no video.
Then repeat with CT60 se to 030 mode, it will most likely start all right.
By the way, what is your 060 CPU revision/mask?
What are the specs of your power brick?

tommo
Atari User
Atari User
Posts: 30
Joined: Mon Jan 29, 2018 6:00 pm

Re: ct60 with a pico psu - startup problematic

Postby tommo » Mon Jan 21, 2019 2:06 pm

Thanks fo the reply,

mzry, The power switch is a 'normal open' type from a old atx-tower.


mpattonm, bootinfo shows: rev1 mask F43g/g65v

Test done:

If i replace the power button (that connects pin 1&2. on the ct) with a bridge so is is set (pwr is grounded/set) then
In 030 and 060 mode when i connect the power brick to the picopsu it goes in off mode(sleep).
The power led is dimly lit(sleep) no ven movement no power led of keyboard.

Forcing the pico with a switch that connect green pin 14 (ps_on) to ground works.
In 030 and 060 mode when i connect the power brick to the pico it goes on.
In this forced way the the ct60 can not switch itself off(sleep) so my little acc "computer off" and the timed shut-down do not work.

Nice knowing it is probably the slow ramp-up of 3.3v, thank you for that.
That would explane that after 2 quick pushes it is more likely to start.

Now if it is posible to make the time the ct60 holds down the green ps-on line it should work. :roll:



Nice you asked for the power specs, so i just measured the new one.

The old way running the ct-falcon
is a 250w (switching) aopen atx that takes about 65w.
the new is a 96w (12v*8A) + the pico what takes only about 30w!

The big tower supply waists more than double !!! 8O

Measure with a conrad (+-5 a 10w ?) device

hencox
Atarian
Atarian
Posts: 9
Joined: Thu Jul 18, 2013 4:34 pm
Location: Göteborg, Sweden
Contact:

Re: ct60 with a pico psu - startup problematic

Postby hencox » Mon Jan 21, 2019 3:29 pm

Hello!

This sounds very similar to my problems. I have a pico-PSU in my falcon also and when I try to power it on it very often just lights up the powerled for a fraction of a second and then dies down again. I've discovered that if I slightly twist the 12V power cable (coming from the AC/DC-adapter and going into the power lead that finally ends in the ATX-connector in the pico-PSU) then it always suddenly comes alive. I think this suggests that I have insufficient contact in that 12V-connector. Maybe you have the same problem?

/Henrik

User avatar
1st1
Atari Super Hero
Atari Super Hero
Posts: 836
Joined: Mon May 07, 2012 11:48 am

Re: ct60 with a pico psu - startup problematic

Postby 1st1 » Mon Jan 21, 2019 9:31 pm

There is a tipp from Chips'b'Chips which can help. Replace C7 capacitor on Falcon mainboard by 10uF, This one delays the reset logic a bit, until the power supply is more stable.
Power without the Price. It's not a bug. It's a feature. _/|\_ATARI

1040STFM in PC-Tower (PAK68/2, OvrScn, 4 MB, 1GB SCSI, CD-ROM...) * 2x Falcon 030 32GB/14MB+ScrnBlstrIII * 2x TT030 73GB/20MB+Nova * 520/1040STFM * 520/1040STE * 260/520ST/+ * some Mega ST * 2x Mega STE 500MB/4MB+M.CoCo * Stacy * STBook * SLM605 * SLM804 * SLM605 * SMM804 * SH 204/205 * Megafile 30/44/60 * SF314 * SF354 * 5x Pofo * PC3

tommo
Atari User
Atari User
Posts: 30
Joined: Mon Jan 29, 2018 6:00 pm

Re: ct60 with a pico psu - startup problematic

Postby tommo » Tue Jan 22, 2019 11:56 am

mpattonm, i see what you mean.
mpattonm wrote:I have meant to say off-spec picoPSU is to blame, specifically slow ramp-up of 3.3 rail
I looked at CT_schematics.pdf. and get it now.

The power switch gives a trigger on the blue line.
This flips the flip-flop green bottom line to the pico-psu to switch the power on or off.
When the purple line goes to 0v (SLEEP) the flip-flop gets a clr and resets (power off).
The purple line is on 5v(sb) trough r22 when shutdown.
and if the ct60 is on it is 3.3v from the SLEEP line of ABE-60 chip.

So no problem,right?

Except when the 3.3v just starts.
It builds up to slow from 0 to 3.3v and the purple line get a fraction of a second to close to zero and the flip-flop gets a CLR and tells the pico to stop the power.

So maybe something can be put over r22 so it will not so easy change state?
You do not have the required permissions to view the files attached to this post.

tommo
Atari User
Atari User
Posts: 30
Joined: Mon Jan 29, 2018 6:00 pm

Re: ct60 with a pico psu - startup problematic

Postby tommo » Tue Jan 22, 2019 12:08 pm

Hi hencox,
Indeed very similar, i tested but the 12v is good.
My power-led brightens not when i insert the 12v, it just goes to dimly lit (SLEEP).

Hi 1st1,
Thank you.
It seems to be that it is 3.3v related, so it must be on the ct60 side.

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Tue Jan 22, 2019 1:25 pm

tommo wrote:So maybe something can be put over r22 so it will not so easy change state?

Indeed it looks as very likely scenario! I think I will pull out scope later on today :). Meanwhile, you could try adding 1-10uF polarized capacitor between IC8 pin 1 and ground, to see if it helps.
Edit: It will only help if that FPGA sleep pin is tri-state. I am not sure about it. Lifting pin 1 of IC8 from PCB is rather safe bet.

User avatar
1st1
Atari Super Hero
Atari Super Hero
Posts: 836
Joined: Mon May 07, 2012 11:48 am

Re: ct60 with a pico psu - startup problematic

Postby 1st1 » Tue Jan 22, 2019 9:02 pm

tommo wrote:Hi hencox,
Hi 1st1,
Thank you.
It seems to be that it is 3.3v related, so it must be on the ct60 side.


I think this is exactly the problem. That capacitor and reset circuit controls the reset signal for mainboard and ct60 CPU. If the CPU starts too early, while power supply is not stable, startup of the Falcon fails. The original Falcon power supply is very good, it gets very quick stable. These pico PSU might be slower. Also my Falcon with ct60e has startup problem, when power on in 060 mode it always fails, in 030 mode it mostly starts. The 060 is powered out of 3,3V supply, the onboard 030 from 5V. One friend in german Atari forum already did that patch with an Afterburner 040 Falcon, and there it helped. I also will try this modification.
Power without the Price. It's not a bug. It's a feature. _/|\_ATARI

1040STFM in PC-Tower (PAK68/2, OvrScn, 4 MB, 1GB SCSI, CD-ROM...) * 2x Falcon 030 32GB/14MB+ScrnBlstrIII * 2x TT030 73GB/20MB+Nova * 520/1040STFM * 520/1040STE * 260/520ST/+ * some Mega ST * 2x Mega STE 500MB/4MB+M.CoCo * Stacy * STBook * SLM605 * SLM804 * SLM605 * SMM804 * SH 204/205 * Megafile 30/44/60 * SF314 * SF354 * 5x Pofo * PC3

tommo
Atari User
Atari User
Posts: 30
Joined: Mon Jan 29, 2018 6:00 pm

Re: ct60 with a pico psu - startup problematic

Postby tommo » Wed Jan 23, 2019 2:08 pm

It was not the CLR/SLEEP that got too low.

I cut the bridge between pin 1 and the round thingy next to it.
so the CLR of the flip-flop would be disabled. (this is very easy to restore with a solder drop)
With the old atx supply as power all was o.k. and indeed i could no longer shut off the falcon with my accessory.
Now with the pico connected nothing changed, so it is not the CLR that got too low.

I also put a 10uF over the C7 to delay the reset logic.
this is in my case also did not do anything.

And the 10uF to stabelise the 5vsb had also no effect. :(



So for now i am using the pico, but i FORCE it ON with a switch on the pico that connects ps_on to ground. The power-switch from the ct60 still works but only to brighten(ON) or dim(SLEEP) the power led in this state.

When the (force)switch is flipped the ct60 takes control back over ps_on and the power stays ON or goes OFF according to the power led. (This is the normal state)


Thanks for the comments so far,
Still open for any suggestion.

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Wed Jan 23, 2019 6:23 pm

There might be other reasons why 7474 keeps resetting. I have not had a time to scope it, but looking at the CT60 schematics, there is no AC filter cap on +5VSB (the always-on supply for 7474) and there is no decoupling cap on 7474 itself either, while I would expect typical 100nF ceramic cap tied to its 7 and 14 pins.
Theoretically, if +5VSB is not stabilized on PicoPSU side as well, 7474 can simply reset on voltage drop during state change. Remember its regular TTL "F" logic, not bit more tolerant CMOS. And I would not be surprised to find out there is no capacitance on that line at all.

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Wed Jan 23, 2019 7:18 pm

Another fix worth to try would be adding higher capacitance in parallel to C36, anything between 1-10uF.

tommo
Atari User
Atari User
Posts: 30
Joined: Mon Jan 29, 2018 6:00 pm

Re: ct60 with a pico psu - startup problematic

Postby tommo » Thu Jan 24, 2019 1:44 pm

If i use the force-switch to power up, the current draw is so great to get all the voltages to normal that the ct power led blinks.
This is with a 12v/8a , a 12v/5a and the atx as 12v power brick.

So the 5vsb goes down with startup. Maybe too far and this indeed shuts itself off when not forced ON.

In ideal condition the ct-powerled should be constant dimly lit and not blink at all because of the constant 5vsb.

Normally you do not notice that because of the brightness change of the led. (dimmed to bright)
With the force-switch the power-led which is in dimmed(sleep) mode stays that way so i notice that.

Seems like the 10uF on pin 14 (+5vsb) and 12 (-) of the 7474 is indeed not enough to stabelize and maybe 100uF might do the trick.
(I now realize i named pin 14 and 12 pin 1 and 3 previously. This is wrong. The first and third pin for top-left are named 14 and 12)

edit:
I put a 100uF on the pico side on the +5vsb to stabelize, led still blinks.

User avatar
calimero
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2291
Joined: Thu Sep 15, 2005 10:01 am
Location: STara Pazova, Serbia
Contact:

Re: ct60 with a pico psu - startup problematic

Postby calimero » Thu Jan 24, 2019 11:42 pm

I have same problem like you: CT60 wont boot from "first attempt" (Falcon030 mode boot every time).
I have PicoPSU, 120W external brick...
and I did not have this problem until last year or two!

Thanks to all for effort to pin down this problem once for all! :)
using Atari since 1986.http://wet.atari.orghttp://milan.kovac.cc/atari/software/ ・ Atari Falcon030/CT63/SV ・ Atari STe ・ Atari Mega4/MegaFile30/SM124 ・ Amiga 1200/PPC ・ Amiga 500 ・ C64 ・ ZX Spectrum ・ RPi ・ MagiC! ・ MiNT 1.18 ・ OS X

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Fri Jan 25, 2019 4:51 pm

tommo wrote:Seems like the 10uF on pin 14 (+5vsb) and 12 (-) of the 7474 is indeed not enough to stabelize and maybe 100uF might do the trick.
(I now realize i named pin 14 and 12 pin 1 and 3 previously. This is wrong. The first and third pin for top-left are named 14 and 12)

7474 chips in DIL, SO(IC) and TSOP packaging have power on pins 7(GND) and 14(VCC). Check the datasheet, if needed: https://assets.nexperia.com/documents/d ... _HCT74.pdf

stormy
Atari Super Hero
Atari Super Hero
Posts: 632
Joined: Tue Jan 26, 2016 12:39 pm

Re: ct60 with a pico psu - startup problematic

Postby stormy » Sat Jan 26, 2019 11:57 am

Lately my pico/ct60e has some odd behaviour too. When I plug in the psu to the pico, the falcon turns on and off for a split second. Then I can normally press the power switch to turn it on.

I worry that this could damage the machine, any thoughts?
Falcon ct60e 060 - 256mb ram - phantom bus and dsp accel // Atari TT - Thunder and Storm IDE 64mb ram - Lightning VME - USB LAN - Ati Mach64 2mb // Atari STM 4160 - 4MB ram - 16Mhz exxos 1.5 booster - Ultrasatan

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Sun Feb 03, 2019 7:25 am

Out of curiosity, I have made some real word testing on my CT60 (in 60 mode):

It does NOT start from PicoPSU80
It does NOT start from brand new Traco power 60316 PSUs
It does MOSTLY start from an old Diametral lab dual output power supply
It ALWAYS starts from no-name 500W ATX PSU

They all work well when 060 is disabled (CT60 in 030 mode)
All of these PSUs are specced well so they ALL should work fawlessly.
And what is most intruding and surprising to me is that it does not start from what is probably one of worlds best PSU available. Traco TOP 60316 has cleanest power output of the mix, it only has 60mV peak-to-peak ripple on 3.3V rail and very fast and clean startup, just 2.4ms. In case of Traco, both main rails also ramp up virtually in the very same moment.
Why, on the other hand, it mostly starts of Diametral PSU, which has about 4 times slower ramp-up (about 10ms) and generates some noise when beeing switched on, is a real mystery to me.

By the way, while doing these tests, I have also measured real consumption:
The CT60 equipped falcon with 512 SDRAM stick in it, 2,5" HDD and floppy drive draws max. 3.1A during startup and idles around 2.3A in desktop on 3.3V rail and peaks at 2.0A/idles at 1.6A on 5V rail.
The above is identical basically identical for 68EC060 rev.1 and 68060 rev.6 CPUs in it.
You do not have the required permissions to view the files attached to this post.
Last edited by mpattonm on Mon Feb 04, 2019 11:30 pm, edited 2 times in total.

User avatar
1st1
Atari Super Hero
Atari Super Hero
Posts: 836
Joined: Mon May 07, 2012 11:48 am

Re: ct60 with a pico psu - startup problematic

Postby 1st1 » Sun Feb 03, 2019 4:50 pm

This is interesting observation, so it depends on power supply. What is the yellow and cyan signal on these diagrams, 5V and 3.3V? I think in 030 mode, the 3,3V power does not care much, but it is used to power the 060 CPU.

Can we also see Falcon's reset signal and "Power Good" Signal of the power supply? With and without bigger C7? Is ct60/ct60e using the Power Good signal?
Power without the Price. It's not a bug. It's a feature. _/|\_ATARI

1040STFM in PC-Tower (PAK68/2, OvrScn, 4 MB, 1GB SCSI, CD-ROM...) * 2x Falcon 030 32GB/14MB+ScrnBlstrIII * 2x TT030 73GB/20MB+Nova * 520/1040STFM * 520/1040STE * 260/520ST/+ * some Mega ST * 2x Mega STE 500MB/4MB+M.CoCo * Stacy * STBook * SLM605 * SLM804 * SLM605 * SMM804 * SH 204/205 * Megafile 30/44/60 * SF314 * SF354 * 5x Pofo * PC3

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Sun Feb 03, 2019 5:46 pm

Yes, lower is 3.3V, upper is 5V.
PWR_GOOD is not connected on/used by CT60 (shame, isnt it?).
As for the C7 / reset line influence: I have already bigger capacity C7 soldered on, I guess its 22uF now. Thats why I can not repeat the tests with original (1uF ?) capacitor. Anyway, I do not see direct relation of power-on issue with Falcon reset circuit. At least in my case,once the CT60 does not start correctly right on power up, such as in test case with Traco TOP 60316, pressing RST switch does not change the situation - it will not reset and kick in, it keeps the black screen.
So its not the same situation several users observed with aftermarket PSUs and drying up capacitors on plain F030, where pressing RST button made seemingly "dead" unit to launch normally.
My wild guess is 3.3V rail has to ramp up (a few ms) earlier than 5V rail in order to initialize FPGAs in time for the bus logic.
Last edited by mpattonm on Tue Feb 05, 2019 6:22 am, edited 1 time in total.

tjam
Atariator
Atariator
Posts: 20
Joined: Sun Dec 09, 2018 10:54 pm
Location: Finland

Re: ct60 with a pico psu - startup problematic

Postby tjam » Mon Feb 04, 2019 8:41 pm

hencox wrote:Hello!

This sounds very similar to my problems. I have a pico-PSU in my falcon also and when I try to power it on it very often just lights up the powerled for a fraction of a second and then dies down again. I've discovered that if I slightly twist the 12V power cable (coming from the AC/DC-adapter and going into the power lead that finally ends in the ATX-connector in the pico-PSU) then it always suddenly comes alive. I think this suggests that I have insufficient contact in that 12V-connector. Maybe you have the same problem?

/Henrik


I have similar problem with my second picoPSU + CT63 falcon combo. Computer boots directly when I connect external power brick to picoPSU, but ATX power switch button does not have any effect. If I twist white 12V power cable at near where it connects to picoPSU, machine resets... If I push white 12V power cable firmly towards to picoPSU, ATX reset switch starts to work (I can power off/on machine), so it looks like problem might be at picoPSU side.

This particular CT63 has really tight ATX connector and I had to loosen up a bit couple of connectors on picoPSU side to make it fit... However, picoPSU stays firmly in CT63 ATX connector.
Last edited by tjam on Mon Feb 04, 2019 10:26 pm, edited 1 time in total.

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Mon Feb 04, 2019 10:12 pm

Added third in the mix to my previous post here.
As you can see Traco has both voltages at nominal levels virtually at the same time, the delta is around 0.1ms.
Diametral ramps up 3.3V rail to nominal some 2ms before 5V rail does.
ATX PSU has the time delta between the two about 3-4ms.
Appearently its what makes the difference.

User avatar
1st1
Atari Super Hero
Atari Super Hero
Posts: 836
Joined: Mon May 07, 2012 11:48 am

Re: ct60 with a pico psu - startup problematic

Postby 1st1 » Mon Feb 04, 2019 11:21 pm

ATX PSU specification says that Power Good signal has to switch from 0V to 5V not sooner as 100 ms after voltage is stable. To me this sounds as it is a good idea to connect the Power Good signal directly to the reset button or the reset signal. Does it make sense to try that?
Power without the Price. It's not a bug. It's a feature. _/|\_ATARI

1040STFM in PC-Tower (PAK68/2, OvrScn, 4 MB, 1GB SCSI, CD-ROM...) * 2x Falcon 030 32GB/14MB+ScrnBlstrIII * 2x TT030 73GB/20MB+Nova * 520/1040STFM * 520/1040STE * 260/520ST/+ * some Mega ST * 2x Mega STE 500MB/4MB+M.CoCo * Stacy * STBook * SLM605 * SLM804 * SLM605 * SMM804 * SH 204/205 * Megafile 30/44/60 * SF314 * SF354 * 5x Pofo * PC3

mpattonm
Hardware Guru
Hardware Guru
Posts: 404
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: ct60 with a pico psu - startup problematic

Postby mpattonm » Tue Feb 05, 2019 10:46 pm

1st1 wrote:ATX PSU specification says that Power Good signal has to switch from 0V to 5V not sooner as 100 ms after voltage is stable. To me this sounds as it is a good idea to connect the Power Good signal directly to the reset button or the reset signal. Does it make sense to try that?

I wish it would be that easy. If I plug a non-working PSU and power on the system while holding a reset button, the system will still not boot up, once the reset switch is released.
In the same test case, performed with working/compliant PSU (such as ATX PSU), the system will boot normally after reset button is released.
Looking at the CT60 schematic I think 060 is not reset by Falcon bus RST signal, but rather by reset a signal generated in one of the FPGA.
I now do not think the PSU compatibility can be improved on physical HW level, I am inclined to say the issue is within FPGAs - either in HDL code, or firmware.
I wonder if Didier or Miro could help.


Social Media

     

Return to “CT60 / CT63 Area”

Who is online

Users browsing this forum: No registered users and 1 guest