i had problems with boost control on my aem ems for quite a while. i was tuning a calibration based off the aem start up map. A tuner i was using stated he has had success with boost control on mk4s but not mk3s.
i found that the base map for mk4 has a map tooth filter of 6 and the mk3 has a zero. so my map readings came off every tooth and not every 6th tooth. my map psi was all over the place as much as 3 to 5 psi oscilating under boost. when i finally found out i need to put a number in the map filter tooth it smoothed out the boost reading and i could program the boost control.
now my boost control is very good. i use a aem boost selinoid and the synapse 50mm wastegate. i use the boost target rpm vs tps table, the boost error duty table, and the boost target comp table. the target table is set to 19.5 psi, my boost control duty table is set to the same duty above 3 psi about 30%. below that it is set much higher to keep the selanoid shut and spoolling. the duty table at 30% gets me a certain psi about 10psi. i also have the boost duty comp table which will add a % amount based on the target psi. so when i put in 19.5 on the boost target rpm vs tps table the aem looks at the boost duty table and gets 30% (as its all set to 30% at all psi) then looks at the boost target comp table at the 19.5 psi and sees a % of 24.61% and adds this to the duty table number of 30% to get 54.61% waste gate duty cycle which with the awesome synapse waste gate (it has no diaphram) the aem selonoid oscilates at 54.61% and the synapse piston oscilates at about 54.61% to obtain 19 to 20 psi.
then if the boost is below target the aem looks at the boost target error duty table and sees that the boost is 10 psi this second and adds the number at -9 psi as the boost is below 19.5. the boost error duty table number at -9 is 44% so the wg duty is maxed out at 90%. then boost rises to 17psi so thats an error of -2.5 and the boost error duty table at -2.5 psi is 5.86 so total duty at this second is 30% off wg duty base table plus boost target comp table at 19.5 of 24.61%, plus boost target error table at -2.5 psi is 5.86 for a total wg duty of about 60%. the error table helps not to over shoot target and keeps the waste gate at target as rpm increases.
when i fllick the boost target switch to 24 psi i have all the same numbers used but the boost target comp table duty % does not use the 19.5 number of 24.61% as now it looks up 24 psi and sees the number i put of 38.28%. the total wg duty output to the selonoid now is 30% of base table plus 38.28 off boost arget comp table for 68.28% wg duty cycle to the aem selonoid. which will result in 24 psi. the boost error duty table will also be added in according to what the actual boost psi is and the target requested plus or minus. It works very well when you know what numbers work the duty table.
the synapse wastegate is awesome. boost is applied to keep it closed and the the selonoid applies presure to open it. there is a large spring on the pistion and a small spring on top of the pistion. the small spring keeps the pistion shut off boost. the large spring keeps the piston up. boost air is applied to both sides of the piston as controled by the seloniod. as there is no diaphram but a piston with air applied to both sides the wg pistion will mimick the operation of the selonoid piston.
once i put in a map tooth filter number of 6 so i could read a steady psi output i tuned the wg in about 30 minutes on the street. you turn off all wg comps and error duty and just use the wg duty map. set it to a number % and boost. so 30% got me about 8psi. 50 % got me about 16 psi, 60 % got 20 psi etc. i then put 30% onthe wg duty table and then put the % duty number in the boost comp duty table according to the above results for that psi target. of course i subtracted out the wg base duty table number of 30 when populating the boost comp duty table. so now when i put in my target boost number the aem used the wg duty map % and the boost comp duty % to get the output at a particular psi as per the above procedure and augments the wg duty out by the error duty amount necessary. figring out the error duty table numbers so they were not too big causing the boost target to over boost and move around or too low causing creep or unsteady psi numbers was the most time consuming parts.
basically once programed it works very well. i do think the synapse wg worked better than the tial 30 mm and the hks 40 mm i used prior as for the design reasons i stated.