Hi there - I've blown up the schematic and gone through the part values piece by piece, so this is quite comprehensive - I am building a number of these, (on my first version at the moment) please comment on your thoughts:
C20 - without it I got 32v and blew the 2506 switch and C10...
the lm2733 datasheet specifies 550pF (confirmed by mightyguru) not the 160pf on ladyada's schematic, and not required on her website. I put 390pF there, and used 12k for R4 and set R3 at 47k as per the datasheet(37k on the website) to get a stable and precise 6v to power the 2506switch, IC8 - lm358 op amp, and the sxa389bs.
C83 and C84 - not on website, but mightyG used them, and they shouldn't be critical - the schematic has them also.
by the way mightyguru - your picture shows the crystal soldered upside down - did you fix it?
R12 - website 68ohm - schematic 100ohm - which is correct?
R27 and R28 - schematic shows 0ohm resitors - a bridge wire? but the website says use ferrites, and mightyguru advised 1206 Ferrite bead = Digikey 240-2413-1-ND - is this OK?
R15 website says 47k, schem says 100k - what is right? VCO1 vtune range is 1-11v (16v max)
R13 website says 39k, schem says 100k - which one? VCO2 tune range is 1-24v (30v max)
Gain stage 2 BIG discrepancies: C63 website says 1uF, but sirenza datasheet for SXA389B and schematic both say 1pF,
C40 schematic and website both say 68pF, but datasheet says 1.2-2.2pF - any suggestions?
smaller gainstage diffs:
sirenza suggests:
18 or 15 nH for L7 - ladyada uses 22nH
C41 should be 47 or 2.2pF, ladyada uses 100pF
C42 should be 22pF, ladyada uses 68pF
C43 should be 47pF or 22pF, ladyada uses 100pF
C48 should be 47pF, ladyada uses 68pF
please comment on the above if you can correct me/point me in the right direction.
Ian
schematic versus website part diffs - full list v1
Moderators: adafruit_support_bill, adafruit
Please be positive and constructive with your questions and comments.
-
- Posts: 102
- Joined: Mon Aug 27, 2007 6:49 am
-
- Posts: 102
- Joined: Mon Aug 27, 2007 6:49 am
by the way the board I'm using is RC1a
that's RC1a
-
- Posts: 63
- Joined: Tue Aug 14, 2007 5:34 am
Re: schematic versus website part diffs - full list v1
Thanks for putting this together! I've been working on a new version of the schematic, but neatened up and with errors corrected.inh wrote:Hi there - I've blown up the schematic and gone through the part values piece by piece, so this is quite comprehensive - I am building a number of these, (on my first version at the moment) please comment on your thoughts:
Comments inline:
I don't really see how that could cause the voltage to skyrocket like that, but I agree that placing C20 is a good idea, and that it should be about 500pF.C20 - without it I got 32v and blew the 2506 switch and C10...
the lm2733 datasheet specifies 550pF (confirmed by mightyguru) not the 160pf on ladyada's schematic, and not required on her website.
Why did you pick 6v? The main users of that supply are the SXA389Bs; their datasheet gives a spec supply voltage of 5v, with an "absolute maximum" of 6v.I put 390pF there, and used 12k for R4 and set R3 at 47k as per the datasheet(37k on the website) to get a stable and precise 6v to power the 2506switch, IC8 - lm358 op amp, and the sxa389bs.
Rys also used them, and so did I; I don't see a reason to leave them out.C83 and C84 - not on website, but mightyG used them, and they shouldn't be critical - the schematic has them also.
The crystal isn't polarized; you can solder it in either way.by the way mightyguru - your picture shows the crystal soldered upside down - did you fix it?
I used the 100ohm from the schematic, but I don't think it is a critical value.R12 - website 68ohm - schematic 100ohm - which is correct?
I'd really like to see these parts removed and replaced with L* symbols so that this confusion disappears. Yes, those will work just fine for R27 and R28.R27 and R28 - schematic shows 0ohm resitors - a bridge wire? but the website says use ferrites, and mightyguru advised 1206 Ferrite bead = Digikey 240-2413-1-ND - is this OK?
... and these are the worst error in the schematic. Those "placeholder" values could damage parts.R15 website says 47k, schem says 100k - what is right? VCO1 vtune range is 1-11v (16v max)
R13 website says 39k, schem says 100k - which one? VCO2 tune range is 1-24v (30v max)
I found these values on the forum:
yopper: R13=56K, R15=27K
fdsa: R13=39K(?), R15=22K
trayvou: R13=47K, R15=75K (but with other component changes)
trayvou: R13=82K, R15=47K (in a different thread)
rys: R13=63K, R15=33K
I used rys's values on mine, but I think the gain is too high -- I'm working on some code to test the tuning range of each VCO, and I'm seeing that I'm tuning maybe 30% beyond the end of what each VCO is capable of, meaning that if I turned down my gain I would see increased accuracy. YMMV.
1pF is correct.Gain stage 2 BIG discrepancies: C63 website says 1uF, but sirenza datasheet for SXA389B and schematic both say 1pF,
I figured these changes were all due to the fact that we're not using the exact frequencies used in the Sirenza reference circuits, but I'm not very good with RF so I will let others comment here.C40 schematic and website both say 68pF, but datasheet says 1.2-2.2pF - any suggestions?
smaller gainstage diffs:
sirenza suggests:
18 or 15 nH for L7 - ladyada uses 22nH
C41 should be 47 or 2.2pF, ladyada uses 100pF
C42 should be 22pF, ladyada uses 68pF
C43 should be 47pF or 22pF, ladyada uses 100pF
C48 should be 47pF, ladyada uses 68pF
There are other, less significant changes I'd like to see made:
* Rename L5 and L6 to reflect the fact that they are really resistors
* Rename R27 and R28 to reflect the fact that they are really ferrite beads
* Change C80 and C81 (the 10MHz crystal caps) from 20 to 22pF, so that they can be combined with an order for the other 22pF caps to save a little bit of money
* Replace C77 with a 1.0uF 0603 cap, since it's really hard to find a 4.7uF 0603 part
* Remove the low-power charging circuitry from the schematic
* Remove the noise generator from the schematic.
I'm sure there must be other things, too, that I'm forgetting.
Also, be aware of the problem with the TPS79333 -- several people have had problems with them randomly burning out. See the "3.3v LDO" thread for more details.
-
- Posts: 102
- Joined: Mon Aug 27, 2007 6:49 am
response to bushing
Thanks for the info bushing. One point about the crystal is that there are 4 pins - top left and bottom right are connected to the can - not in use (looking at the writing on the can from above the right way). bottom left and top right are connected to the crystal. This means that if you mount the crystal upside down as in MG's picture - the inactive can leads will be connected to the PLL pads, and the crystal will be connected to the 2 spare pads on the board which don't appear to be connected to the PLL. please correct me if I have missed something.
re 6v - I guess that I saw 6v specified on the schematic, and figured that it might be reduced a little by the time it got to the sxa389bs...
I think the lm2733 output voltage may have skyrocketed by having oscillator delay time nixed by having no cap to "take the cycle through zero" - but this is gut feeling... I thought the oscillator was a fixed 1.6MHz fq, so maybe there is another explanation - buck/boost converters are not my strong point...
when you say the low power charging circuitry - what do you mean? - I'm not sure I see anything except the 4 power supplies...
last thing - what is the point of the 4 pin jumpers - it seems that GND and +3.7v power gets transmitted via the metal spacers?
Do the middle 2 pins connected to pins 30 and 31 of the atmega 88 do something?
thanks
Ian
re 6v - I guess that I saw 6v specified on the schematic, and figured that it might be reduced a little by the time it got to the sxa389bs...
I think the lm2733 output voltage may have skyrocketed by having oscillator delay time nixed by having no cap to "take the cycle through zero" - but this is gut feeling... I thought the oscillator was a fixed 1.6MHz fq, so maybe there is another explanation - buck/boost converters are not my strong point...
when you say the low power charging circuitry - what do you mean? - I'm not sure I see anything except the 4 power supplies...
last thing - what is the point of the 4 pin jumpers - it seems that GND and +3.7v power gets transmitted via the metal spacers?
Do the middle 2 pins connected to pins 30 and 31 of the atmega 88 do something?
thanks
Ian
-
- Posts: 63
- Joined: Tue Aug 14, 2007 5:34 am
Re: response to bushing
(BTW, these are all great questions. LadyAda, if you're reading this, would you be opposed to making the "Make" pages for the WB into a Wiki, so that we could update them appropriately? thanks.)
1.............3
2.............4
1 and 4 are the two pins connected to the crystal; 2 and three are grounded to the shield.
If you rotate the chip 180 degrees, you get
4.............2
3.............1
So, 1 is where 4 was, and 4 is where 1 was, so it still works. right?
That having been said -- if you look at the SXA datasheet, it mentions two voltages: +5v supply, absolute maximum 6v. They don't say what happens in between; maybe we'll get better performance out of the amp if crank the power up to 6v? (At the expense of part longetivity, maybe?)
http://www.ladyada.net/make/wavebubble/battery.html
then scroll all the way down to the bottom -- it's the second note from the bottom.
-b
It's hard to visualize, but try this. Say, the crystal as four pins:inh wrote:Thanks for the info bushing. One point about the crystal is that there are 4 pins - top left and bottom right are connected to the can - not in use (looking at the writing on the can from above the right way). bottom left and top right are connected to the crystal. This means that if you mount the crystal upside down as in MG's picture - the inactive can leads will be connected to the PLL pads, and the crystal will be connected to the 2 spare pads on the board which don't appear to be connected to the PLL. please correct me if I have missed something.
1.............3
2.............4
1 and 4 are the two pins connected to the crystal; 2 and three are grounded to the shield.
If you rotate the chip 180 degrees, you get
4.............2
3.............1
So, 1 is where 4 was, and 4 is where 1 was, so it still works. right?
I just checked, and the switches are rated at resistances in the milliohm range, so I would be surprised to see a measurable voltage drop.re 6v - I guess that I saw 6v specified on the schematic, and figured that it might be reduced a little by the time it got to the sxa389bs...
That having been said -- if you look at the SXA datasheet, it mentions two voltages: +5v supply, absolute maximum 6v. They don't say what happens in between; maybe we'll get better performance out of the amp if crank the power up to 6v? (At the expense of part longetivity, maybe?)
These power supplies are like voodoo to me, too -- all I can say is that I originally built five boards without any C20 and they all seemed to work fine, but that I also agree it should be placed "just in case".I think the lm2733 output voltage may have skyrocketed by having oscillator delay time nixed by having no cap to "take the cycle through zero" - but this is gut feeling... I thought the oscillator was a fixed 1.6MHz fq, so maybe there is another explanation - buck/boost converters are not my strong point...
It's part of the LiPoly board.when you say the low power charging circuitry - what do you mean? - I'm not sure I see anything except the 4 power supplies...
http://www.ladyada.net/make/wavebubble/battery.html
then scroll all the way down to the bottom -- it's the second note from the bottom.
I also don't understand why we're passing power both through 2 of the mounting spacers and through two of the pins of that 4-pin header -- but the middle two pins carry serial data from the ATMega to the FTDI chip so you can interact with the device to configure it after / while you are building it.last thing - what is the point of the 4 pin jumpers - it seems that GND and +3.7v power gets transmitted via the metal spacers?
Do the middle 2 pins connected to pins 30 and 31 of the atmega 88 do something?
-b
-
- Posts: 102
- Joined: Mon Aug 27, 2007 6:49 am
crystal - you're right of course
yes of course you are right about the crystal, good thing I'm not an airplane pilot...
- dawsrobe
- Posts: 7
- Joined: Sun Apr 06, 2008 11:14 am
Re: schematic versus website part diffs - full list v1
Hey Scoopin:
Did you ever come up with the code for testing the tuning range of the VCO's?
I've used values from the forum, but I would like to verify for myself.
Anything you can do would be appreciated.
Did you ever come up with the code for testing the tuning range of the VCO's?
I've used values from the forum, but I would like to verify for myself.
Anything you can do would be appreciated.
- dawsrobe
- Posts: 7
- Joined: Sun Apr 06, 2008 11:14 am
Re: schematic versus website part diffs - full list v1
I guess I should have mentioned inh in this too. You've been a big help in the past.
I'm trying to adjust the Vtune max values, but I think I need to adjust code to do this. How exactly did you go about measuring your Vtune max values?
Hope you can help.
Thanks.
I'm trying to adjust the Vtune max values, but I think I need to adjust code to do this. How exactly did you go about measuring your Vtune max values?
Hope you can help.
Thanks.
- aqisnotliquid
- Posts: 117
- Joined: Mon Oct 06, 2008 10:46 pm
Re: schematic versus website part diffs - full list v1
bushing, You know, a wiki would be a fantastic idea. In fact, I think she already beat us to it.
http://ladyada.net/wiki/wavebubble/start
I'll be putting information into here when I get the chance. The only things really keeping me from doing it are school and work, but in between, I'll see about putting things up.
Right now, I'm working on a comprehensive parts list. Although, once I posted it, I found out that I wasn't the only one that had wanted such a thing, and antares posted one of his own. His was much more organized, so I thought I'd go ahead and put it up on the wiki.
http://forums.ladyada.net/viewtopic.php?f=16&t=7416
There's the link to both my list and his list. Mine is purely off the site. His is every part, but only from Digikey.
It would be amazing if we could work together and get a parts list that everyone can use. That way, we can cut down on the majority of the confusion in this board about what parts to use and start threads about improvements.
http://ladyada.net/wiki/wavebubble/start
I'll be putting information into here when I get the chance. The only things really keeping me from doing it are school and work, but in between, I'll see about putting things up.
Right now, I'm working on a comprehensive parts list. Although, once I posted it, I found out that I wasn't the only one that had wanted such a thing, and antares posted one of his own. His was much more organized, so I thought I'd go ahead and put it up on the wiki.
http://forums.ladyada.net/viewtopic.php?f=16&t=7416
There's the link to both my list and his list. Mine is purely off the site. His is every part, but only from Digikey.
It would be amazing if we could work together and get a parts list that everyone can use. That way, we can cut down on the majority of the confusion in this board about what parts to use and start threads about improvements.
-
- Posts: 94
- Joined: Tue Jul 31, 2007 2:52 am
Re: schematic versus website part diffs - full list v1
the best way to test a vco is make a test bored apply vcc to the vco what ever the vco voltage calls for . then use a 10k 10 turn pot to the vt of the vco and you need to pad down the rf out of the vco. use a 27ohm in series at the rf out then at the end of the 27ohm use a 33ohm to ground then at the two points meet use a 150 ohm in series . thats your rf out if you dont do this the vco will drift every where and the test will be invalid . and you need a spectrum analyzer to see how far it goes in the band....dawsrobe wrote:Hey Scoopin:
Did you ever come up with the code for testing the tuning range of the VCO's?
I've used values from the forum, but I would like to verify for myself.
Anything you can do would be appreciated.
-
- Posts: 18
- Joined: Thu Oct 16, 2008 3:34 am
Re: schematic versus website part diffs - full list v1
thanks bushing
you always have some great info and output on this complex project
you always have some great info and output on this complex project
- sheepke
- Posts: 8
- Joined: Fri Dec 26, 2008 6:14 pm
Please be positive and constructive with your questions and comments.