Pure Evoke H4 H6 DAB radio No power, sound, no menu scrolling, Waiting-PC Wizard

[rh100605]

The Pure H series radio seem to have three common faults. Fixing these is not for the faint hearted as you will need a good 5x magnifier a hot air smd solder gun and a powerful soldering iron to remove screening cans.  Then some very careful use of flux and solder wick with a needle point iron. So I used three soldering irons...…

First problem
Sound is available at the headphone output but not from the speakers. The headphones have their own amplifier (ETK 4812).  I used a scope to look at the differential input pins ( 2 x 2 pair) to the  Yamaha YDA 147-S class D stereo amplifier chip  and audio was present as was 12 volts. The Mute pin was at a high level and responded to putting a jack plug into the headphone socket. The YDS 147 is a 48 pin flat pack with a soldered heat sink underneath used on the Pure H4/6 and D4/6. To gain access I removed one wall of the screening can and the four ferrite chokes.

I bought a YDA 147-s on Ebay and used the CHIP QUIK SMD1NL KIT solder (lowers solder melting point) and flux to flood the pins and then a hot air solder removal tool to remove the entire chip. The connection to the mute pin damaged easily but was repairable due to a pad on the pcb. Replacement after lots of cleaning and new flux was by heating by hot air to solder the heat sink and then flooding the pins with solder and removal again with solder wick, very carefully with a needle point iron.

Second problem
Will not power up due to short on 12.6 volt line inside power amplifier can. Normally due to 0805 capacitor short near device marked BA5 but can be the TI Boost buck regulator

Third problem
Some radios start up with "Waiting for PC wizard" and will do nothing else. Others will not scroll through menus. Both are the same area .
Note: this can be caused by powering up the radio with the select push knob depressed.

If removing all sources of power including the battery does not clear the message then it is  the way the processor senses buttons at fault.

The KINO 4 chip inside the smaller can has 2 differential A to D inputs which are connected to two resistor chains  connected to 1.88 volts through 47k. Pushing a button on the front panel grounds a point in one of the chains. You can probe these chains on CN2 on the front panel as pins 1 and 2  ( 3 is the power switch 7 is 3.3v and 8 ground other are volume and light sensor). With no button pushed pins 1 and 2 should be about 1.7 volts or greater. The scroll is at the bottom of both chains so the voltage change is very small to about 1.6volts. Mute and Select push controls are at the top so ground the chain easily .
The problem is that any high impedance short at the chip inputs cause the top of chain voltage to drop so scroll never reaches 1.6 volts. A near short results in the software upgrade command being activated.

Chain 1 (pin1) is Mute, PR1, 2 3 4 5,6+, scroll R.  Chain2 (pin2) is Select, timer, alarm, source, back, menu, N/c, scroll L. About 200mv for each step and approximately. 200k total each
 
I removed the can over Kino4 chip and examined the corner nearest the end of the board and towards the centre. There are four 0 ohm 0402 resistors in a row these are the A to D inputs. The pins on the Kino chip looked unclean so put no clean flux on and used a hot air gun to heat up the chip on the corner until I could see reflow. and voila... I also think the capacitor on the inputs could also leak enough to cause a problem as the chip input impedance is 2M ohm or more. Any fine solder hairs will cause a problem.


Best of luck

[rh100605]

Update Nov 2021

I have fixed two H6 radios that did not power up. I suspected the audio power amplifier so took off the can lid and searched for 12volts  which was available at the ferrites on the input to the can but not on the Yamaha audio amplifier when powered by the mains adapter.
The Pure radio Taymar board is common to H4 and H6 with only the H6 populated for stereo.

Inside the can are 3 regulators. A 12.6 volt synchronous boost regulator (Texas instruments TPS61088RHLR, the input can be 12volt or 4.2v from battery) which supplies 12.6 volts to 2 other switching regulators as well as the class D audio amplifier through zero ohm resistors.
I next measured the resistance of the 12volt input to the amplifier to ground ( 3.2 ohms !).

By removing each zero ohm resistor until the short disappeared, I isolated the 4.6 volt regulator marked BA5 (Torex XC9248) as the source of 3 ohms to gnd. I removed the chip only to find that the short persisted and the only components left in circuit were the bypass capacitors. Two 0805 , one 0603 and an electrolytic . I found the 0805 capacitor furthest from the BA5 chip to be at fault in both radios!

I replaced the faulty capacitor with a 2.2uf 30v murata 0805 capacitor all is now well

[timmsel]

Jo, thanks for your helpfull description. My Radio has the Same Problem (At the powerplug there are only 1.5v and about 2 Ohms to ground at the two caps of the Pin1 Input of the BA5.)
 I removed C407 (he allready has a little scratch on the top , showing on the Picture)
After That the Short is gone but the Radio is still dead and i habe to Go further and wait for the 2.2uF 50V cap i ordered. Hope to Get it working again.

Regards from Kiel

[rh100605]

Kiel
In your post you did not say whether the 12.6 volt rail that supplies the BA5 chip was working. If not it is the Texas instruments boost converter that has failed (TPS61088). If the input to the boost converter is not 12V then check the 8 pin Pch FET FDS44358Z  it should pass 12v from the power input to the boost converter.
If 12.6 is present check the output of the BA5 is 4.6v this then feeds a Torex USP-4  3.3v Regulator XC6221 (3B2) then the Tone control chip ET2354 which can be isolated by a 0 ohm resistor.
I have also found another 0805 capacitor that went short  after the 3.3v regulator (C509) I replaced it with a 1uf Murata part. Also check the bypass capacitors on the ETK 4812 Headphone amplifier.

If the short is present at the input and output of the TPS61088 switching regulator then remove all the adhesive holding the 3 electrolytic capacitors to the pcb ( I used a very small screwdriver and twisted the adhesive onto the screwdriver) and remove the 2 capacitors on the input and stand up the 1000uF capacitor on the output. This prevents damaging them with hot air.

Then remove the regulator using hot air below then above.

If the short goes away test the radio by placing a link between the input and output +12volt capacitor terminals.
The radio should work on the mains adapter (not battery)

[rh100605]

Kiel
In your post you did not say whether the 12.6 volt rail that supplies the BA5 chip was working. If not it is the Texas instruments boost converter that has failed (TPS61088). If the input to the boost converter is not 12V then check the 8 pin Pch FET FDS44358Z  it should pass 12v from the power input to the boost converter.
If 12.6 is present check the output of the BA5 is 4.6v this then feeds a Torex USP-4  3.3v Regulator XC6221 (3B2) then the Tone control chip ET2354 which can be isolated by a 0 ohm resistor.
I have also found another 0805 capacitor that went short  after the 3.3v regulator (C509) I replaced it with a 1uf Murata part. Also check the bypass capacitors on the ETK 4812 Headphone amplifier.

I have recently found that an 0805 capacitor ( 1uf?) on the front panel that decouples the 3v3 supply went short. Try removing the display plug and see if the 3v3 supplies on the board recovers.